Session 325 Clinical Imaging

Transcription

ARVO 2016 Annual Meeting Abstracts
325 Clinical Imaging - Miscellaneous
Tuesday, May 03, 2016 8:30 AM–10:15 AM
Exhibit/Poster Hall Poster Session
Program #/Board # Range: 3371–3414/D0308–D0351
Organizing Section: Multidisciplinary Ophthalmic Imaging Group
Contributing Section(s): Lens, Low Vision
Program Number: 3371 Poster Board Number: D0308
Presentation Time: 8:30 AM–10:15 AM
Window to the central nervous system–Retinal examination for
early diagnosis of Alzheimer’s disease
Lily Yu-Li Chang1, 2, Monica L. Acosta1, 2, Joanna Black1, 2.
1
Optometry, University of Auckland, Auckland, New Zealand; 2New
Zealand National Eye Centre, The University of Auckland, Auckland,
New Zealand.
Purpose: Alzheimer’s disease (AD) is a public health priority due
to population aging. The diagnosis is often delayed, and is the major
obstacle to therapeutic success. The eye has attracted much interest
for effective detection of AD due to distinctive visual symptoms in
early stages of AD. Recent studies have shown proteins implicated
in AD may be identified in vitro, and more recently, in vivo, through
retinal imaging. We therefore hypothesized that visual disturbances
in AD may be attributed to changes in the retina, and that there are
differences in retinal structure and function between healthy controls
and AD patients, detectable byan eye examination.
Methods: Retinal structure and function were assessed in 3 group
of people in a pilot study. 32 participants were enrolled as young
controls(age 18-59,N=20), elderly controls(age≥60,N=8), and
dementia patients(age≥60,N=4) including AD(N=2), cerebral amyloid
angiopathy(N=1) and vascular dementia(N=1). Retinal structure was
assessed by fundoscopy, and optical coherence tomography (OCT).
Retinal ganglion cell and macula function were assessed by pattern
electroretinogram (PERG) and multifocal ERG (MfERG). MannWhitney U test and 95% confidence interval was used for statistical
analysis.
Results: OCT showed no significant difference between young and
elderly controls. There was a significant difference in PERG latency,
MfERG amplitude and latency (p<0.05) between the young and
elderly controls. This suggested an aging effect in retinal function
but not in retinal structure. All dementia patients had retinal thinning
in the temporal quadrant(s) of the optic nerve, and in the macula
compared to elderly controls. This structural thinning was mirrored
by delayed latency and amplitude in both PERG and MfERG. AD
cases had the greatest retinal thinning and the most delayed MfERG
latency compared to other dementia.
Conclusions: These findings provide evidence that there is a decline
in retinal structure and function in AD. We speculate this extent of
decline results from a chronic process, and could have started at
pre-dementia stage. The decline of retinal integrity in AD also shows
distinctive difference from other dementia. Retinal examination
requires minimal cognitive input and is easy to perform in clinic. This
investigation suggests that retinal examinations combined with other
ocular tests has the potential to be employed as clinical markers for
AD.
Commercial Relationships: Lily Yu-Li Chang, None;
Monica L. Acosta, None; Joanna Black, None
Retinal Biomarkers of Alzheimer’s Disease
Heather Whitson1, 2, Sina Farsiu3, Sandra S. Stinnett3, Leon Kwark3,
Guy Potter4, James Burke5, Scott W. Cousins3, Eleonora M. Lad3.
1
Medicine, Duke University, Durham, NC; 2GRECC, Durham VA
Medical Center, Durham, NC; 3Ophthalmology, Duke University,
Durham, NC; 4Psychiatry and Behavioral Sciences, Duke University,
Durham, NC; 5Neurology, Duke University, Durham, NC.
Purpose: Retinas of people with Alzheimer’s Disease (AD)
may experience neuroinflammatory processes, similar to brain.
Inflammatory injury may cause atrophy of axonal projections and/
or extracellular deposits. Our objective is to determine whether
nerve fiber layer (NFL) thickness or peripheral (non-macular) drusen
deposits may serve as non-invasive, inexpensive biomarkers to help
diagnose AD.
Methods: NCT01937221 enrolled three age-matched groups: 15
patients with mild cognitive impairment (MCI)/prodromal AD, 15
patients with moderate AD, 17 cognitively normal adults. Assignment
to group was made by consensus diagnosis of a neurologist and
neuropsychologist. We excluded eyes with major eye diseases
or diagnoses that may cause NFL thinning (e.g. normal tension
glaucoma). Study participants underwent examination, spectral
domain optical coherence tomography (SD-OCT), wide-field
fundus color and autofluorescence (AF) photography and stereo disc
photography. Location-specific NFL thicknesses were measured using
Duke Optical Coherence Tomography Retinal Analysis Program
(DOCTRAP) software. Mean NFL thicknesses are compared with
generalized estimating equations (GEE). Peripheral drusen were
graded by retinal specialists masked to group assignment. Proportion
of participants with evident drusen was compared by chi-square test.
Participants are returning for 1-year follow-up exams.
Results: Participants in the AD group were more likely to have
peripheral drusen visible on color or AF photography (66.7%),
compared to participants in MCI (35.7%) or control (37.5%) group
(p=0.06). When drusen was present, it was bilateral in 16 of 17
subjects with two eligible eyes. Preliminary analysis from macular
OCT did not reveal significant group differences in macular NFL
thickness. Ongoing work will address artifact (e.g., ERMs) and
evaluate NFL/GCL thickness around the optic nerve. We will present
longitudinal analyses to address stability/progression of biomarkers.
Conclusions: Compared to age-matched controls or subjects
with MCI, patients with moderate AD had a higher burden of
peripheral retinal drusen deposits. Further analysis will pinpoint
location of deposits within retinal layers and evaluate other means
of quantification. Our preliminary negative findings with regard
to macular NFL thickness controvert a previous group’s finding of
thinner NFL in persons with moderate-to-severe AD. We will present
ongoing analysis on other measures of retinal thickness.
Commercial Relationships: Heather Whitson, None;
Sina Farsiu, None; Sandra S. Stinnett, None; Leon Kwark, None;
Guy Potter, None; James Burke, None; Scott W. Cousins, None;
Eleonora M. Lad, None
Support: Alzheimer's Association NIRG-13-282202; Duke Institute
for Brain Sciences; NIH P30 EY-005722
These abstracts are licensed under a Creative Commons Attribution-NonCommercial-No Derivatives 4.0 International License. Go to http://iovs.arvojournals.org/
to access the versions of record.
Retinal imaging in early and late Alzheimer’s disease
Lajos Csincsik1, Timothy Shakespeare2, Nicola Quinn4,
Ruth E. Hogg4, Sebastian Crutch2, Ritchie Craigh5, Tunde Peto3, 1,
Imre Lengyel1. 1ORBIT, UCL Institute of Ophthalmology, London,
United Kingdom; 2Dementia Research Centre, UCL Institute of
Neurology, London, United Kingdom; 3NIHR Biomedical Research
Centre, Moorfields Eye Hospital NHS Foundation Trust, London,
United Kingdom; 4Queen’s University - Centre for Experimental
Medicine, Belfast, United Kingdom; 5University of Edinburgh,
Edinburgh, United Kingdom.
Purpose: It has been suggested that the development of Alzheimer’s
disease (AD) is associated with thinning of the peripapillary
retinal nerve fiber layer (ppRNFL) due to ganglion cell loss and
accumulation of drusen and that these reflect atrophy and plaque
deposition in the brain, respectively. Here we report the results of
baseline and progression characteristics of two prospective small
studies that examined changes on ultra-widefield imaging (UWFI)
and spectral-domain optical coherence tomography (OCT) to gain
insights into retinal changes associated with AD.
Methods: Color and autofluorescence UWFI and OCT images
were acquired by Optos 200TX and Optos OCT SLO, respectively.
In study 1 (S1), UWFI images of 72 healthy controls (HC) and 46
AD patients (Mini Mental Score Examination (MMSE) <20) were
analysed for presence/absence and progression of drusen at baseline
and at 2-year follow-up. In study 2 (S2), ppRNFL and peripapillary
whole retinal (ppWR) thickness as well as macular volume (MV)
and thickness (MT) were analysed at baseline and at 1-year followup (FU) on OCT images of 72 [4 FU] HC (MMSE>28), 23 [3 FU]
typical AD (tAD) (MMSE<20) and 26 [5 FU] posterior cortical
atrophy (PCA; MMSE<22) patients. The studies had full Ethical
Committee approval. Statistical analysis was carried out using
STATA and SPSS.
Results: In S1, using UWFI we found that there is a significantly
higher prevalence of hard drusen in the peipheral retina only in AD
patients at baseline (p<0.05). Drusen numbers increased significantly
in AD patients by the 2-year follow-up (p<0.05). In S2, using OCT,
disease duration at baseline was not associated with any of the
examined parameters (p>0.05 for all groups). However, the ppRNFL
thickness measurements showed a significant age related decline
(p<0.006). Importantly, UWFI and OCT imaging of the retina was
quick and well tolerated in both studies. There were some difficulties
with maintaining fixation or generating good quality images of
patients with more advanced disease.
Conclusions: These studies suggest that peripheral retinal hard
drusen deposition is associated with AD, suggesting that UWFI
might become a useful tool in monitoring progression in AD. The
clinical utility of OCT on individual patient’s progression is yet to be
determined.
Commercial Relationships: Lajos Csincsik;
Timothy Shakespeare; Nicola Quinn, None; Ruth E. Hogg, None;
Sebastian Crutch, None; Ritchie Craigh, None; Tunde Peto, None;
Imre Lengyel, None
Support: OPTOS plc
A technique for measurement of ganglion cell and retinal
nerve fiber layer thickness change with age in an Alzheimer’s
disease (AD) mouse model using human spectral domain optical
coherence tomography (OCT)
Keegan Harkins, Mohammad A. Sadiq, Vikas Gulati,
Shane J. Havens, Shan Fan, Tara Rudebush, Deepta A. Ghate.
Ophthalmology, Truhlsen Eye Institute - UNMC, Omaha, NE.
Purpose: To determine a technique of measuring mouse ganglion
cell layer (GCL) + retinal nerve fiber layer (RNFL) thickness with
a human OCT machine, to establish a grading protocol for animal
studies that was as stringent as human studies, and to look for
differences in the GCL + RNFL in young and old Alzheimer mouse
models.
Methods: The study population consisted of 9 APPswe, PSEN1dE9
mice including 3 young mice (average age 10±3 weeks) and 6
old mice (average age 41±2 weeks). Spectralis OCT® machine
fitted with a +25 lens with a reference arm adjustment was used
to acquire 12 radial optic nerve head images of one dilated eye in
the anesthetized mouse. The optic nerve head was separated into
3 sectors (12-2 O clock, 2-4 O clock and 4-6 O clock). The expert
grader selected 1 of 4 images from each sector based on image
quality. Two primary graders measured GCL + RNFL thickness 500
µm left and right of the bergmeister papillae and internal limiting
membrane intersection on the optic nerve head (figure 1) from
each image. The mean value of the 2 graders measurements were
used. The grading was evaluated by an independent observer. Any
discrepancies >20 µm were re-graded by the expert grader. If the
expert grader’s measurements were within 20 microns of one of the
graders, the median value of the measurement was used. All graders
were masked to each other.
Results: Nine eyes for a total of 27 radial OCT images were graded.
Of the 54 measurements, only 9 had a discrepancy greater than 20
microns (and needed the 3rd grader) and only 2 had a discrepancy
greater than 30 microns. Of the 54 measurements, only 2 were
classified as ungradable. The intraclass correlation coefficient was
0.7607 between the two graders. The mean GCL + RNFL thickness
was 84.8 µm (SD 1.8 µm) in young mice and 89.0 µm (SD 4.2 µm)
in old mice.
Conclusions: Our study demonstrated that it is possible to acquire
OCT images of good quality in rodents using a human OCT machine.
Our grading protocol was validated by the excellent concordance
between graders. OCT studies in animals should follow grading
protocols that are as strict as human grading protocols. The mean
thickness of GCL + RNFL did not seem to negatively correlate with
increasing age in the AD mouse models.
Commercial Relationships: Keegan Harkins, None;
Mohammad A. Sadiq; Vikas Gulati, None; Shane J. Havens,
None; Shan Fan, None; Tara Rudebush, None; Deepta A. Ghate,
None
Support: American glaucoma Society and Research to prevent
blindness
Assessment of corneal properties using Optical Coherence
Tomography Speckle: A pilot study
Danilo A. Jesus, D Robert Iskander. Wroclaw University of
Technology, Wroclaw, Poland.
Purpose: To explore the potential of Optical Coherence Tomography
(OCT) speckle to assess biomechanical properties of cornea and to
track their changes.
Methods: Several statistical models were used to fit the corneal
backscatter OCT data. Parameters of competing models were
calculated using maximum likelihood estimation and the best
model was selected based on the goodness of fit. The influence of
choosing different regions of interest was also taken into account.
The applicability of the best model was tested analyzing corneal
speckle statistics of subjects with different ages. In addition, variation
of the corneal backscattering on subjects wearing contact lenses was
tracked.
Results: Generalized Gamma Distribution has proven so far, to be the
best model to fit the OCT corneal speckle. Its scaling and two shape
parameters have shown to be sensitive to the variation of corneal
backscattering properties. Differences among people with different
ages were observed, showing the possibility to access corneal
age-related changes using OCT corneal envelope statistics. Similar
results were obtained for subjects wearing contact lenses showing
that corneal backscattering statistics can also be used to track diurnal
changes of corneal biomechanical properties.
Conclusions: This study shows that Generalized Gamma distribution
can be used to fit the corneal speckle and has a great potential to be a
helpful complement to evaluate the structure and elastic properties of
the cornea in-vivo.
(a) OCT image of the cornea where the alternated red and green
rectangles denote different regions of interest and (b) their respective
probability density functions using Generalized Gamma distribution.
(c) Corneal OCT image of a subject wearing a contact lens where the
region of interest is denoted by the red line and (d) the corresponding
envelope of the speckle data and its fit using Generalized Gamma,
Gamma, Rayleigh and Nakagami distributions.
Commercial Relationships: Danilo A. Jesus; D Robert Iskander,
None
Support: Marie Curie ITN grant, AGEYE, 608049.
Microscope-integrated Intraoperative Optical Coherence
Tomography Enables Real-Time Monitoring Of Corneal
Structural Alterations During Collagen Crosslinking
Sebastian E. Siebelmann1, 2, Jens Horstmann1, 2,
Manuel M. Hermann1, Björn Bachmann1, Claus Cursiefen1,
Philipp Steven1, 2. 1Department of Ophthalmology, University of
Cologne, Cologne, Germany; 2Cluster of Excellence: Cellular Stress
Responses in Aging-associated Diseases (CECAD), University
Hospital of Cologne, Cologne, Germany.
Purpose: Corneal UVA-riboflavin collagen crosslinking (CXL) is
a standardized and safe procedure for the treatment of keratoconus.
The main complication is corneal decompensation due to endothelial
cell damage after UVA irradiation. Standardized protocols prevent
irradiation damage in most cases, but no intraoperative parameters
exist for individual procedure surveillance e.g. depth of riboflavin
penetration or progression in collagen crosslinking. Nonetheless
the development of a hyperreflective zone in optical coherence
tomography (OCT) imaging immediately after CXL is assumed
to reflect riboflavin penetration depth. Here we demonstrate the
real time acquisition of changes in corneal structural properties by
intraoperative OCT (iOCT).
Methods: Prospective case series of seven patients (4 female, 3
male), who underwent CXL, using a commercially available 840nm
Spectral Domain OCT (iOCT; OpMedT, Luebeck, Germany), adapted
to an operating microscope (HS Hi-R Neo 900A, Haag Streit, Wedel,
Germany). Underlying disease in all cases was Keratoconus. Increase
of central and peripheral depth-dependent corneal hyperreflectance
was observed in cross-sectional images of the cornea, selected from
volume scans, which were acquired after corneal abrasion, after 30
minutes of riboflavin application and 15 and 30 minutes of UVA
irradiation. For image analysis, a custom-made algorithm using
Matlab-Environment evaluated time-dependent local changes in
corneal reflectivity. Images taken prior to the treatment served as
controls.
Results: In all patients, a hyperreflective zone was visible during
OCT imaging after riboflavin application and after 15 and 30 minutes
of UVA irradiation. Depth and hyperreflectance, were progressive
during the procedure. Reflectance was more intense in the corneal
periphery then in the center.
Conclusions: iOCT enables online imaging and procedure
monitoring during CXL. As the depth of a clearly visible and
well evaluable hyperreflective zone while CXL is associated with
riboflavin penetration depth, iOCT allows for individual penetration
surveillance in real time. Nonetheless it has to be clarified, how
the hyperreflective corneal zone allows for drawing conclusions
regarding e.g. rigidity and arrangement of corneal collagen fibers. In
the future, intraoperative OCT may allow for developing individual
treatment protocols for CXL.
Commercial Relationships: Sebastian E. Siebelmann;
Jens Horstmann, None; Manuel M. Hermann, None;
Björn Bachmann, None; Claus Cursiefen, None; Philipp Steven,
Haag Streit Surgical (R)
Support: DFG FOR 2240 “(Lymph)Angiogenesis And Cellular
Immunity In Inflammatory Diseases Of The Eye”, www.for2240.de;
EU COST BM 1302
A New Method for Measuring Corneal Epithelial Thickness
Esther Young, Homayoun Bagherinia, Patricia Sha, Mary K. Durbin.
R&D, Carl Zeiss Meditec, Inc., Dublin, CA.
Purpose: Evaluating the thickness of the epithelial layer of the
cornea can aid in detection of keratoconus (KCN). We reviewed a
prospective, multi-site study to develop a new method for measuring
corneal epithelial thickness in normal eyes and eyes with keratoconus
imaged using CIRRUS OCT.
Methods: The data used in this study was retrospectively reviewed
from a prospective multi-site verification study. 16 eyes without
ocular pathology (“normal”) and 16 eyes with keratoconus were
imaged with the Pachymetry scan three times on each of six CIRRUS
OCT devices (ZEISS, Dublin, CA). The algorithm identifies the
anterior corneal surface and the upper boundary of the Bowman’s
layer in each 2-D B-scan (24 radial B-scans over 6 mm) using a
combined graph theory and dynamic programming framework.
The epithelial thickness map is created as the closest distance from
each anterior surface point to the Bowman’s layer followed by
interpolation to a complete 2-D map. All images were qualified
by manual review to ensure the algorithm properly marked the
boundaries of the epithelium. Summary parameters were created by
calculating the mean, minimum, and maximum in each hemisphere
divided by the vertex. Analysis of variance was used to identify
the reproducibility standard deviation and the coefficient of
reproducibility was calculated by dividing this by the mean.
Results: Examples of epithelial thickness maps from three
instruments for one normal eye and two keratoconic eyes are
shown in Figure 1. Due to the importance of data quality for this
algorithm and the severe pathology present in some eyes, 20-30% of
images were excluded for poor quality or algorithm failures. Three
keratoconic eyes were excluded due to variability that occurred from
surgical procedures (i.e. keratoplasty, Intacs). Three normal eyes
had algorithm failures that prevented statistical analysis, and were
excluded as well. Of the rest, the mean and standard deviation for
each metric is shown in Table 1.
Conclusions: High-resolution SD-OCT is able to map the epithelial
thickness with excellent reproducibility. The epithelial thickness
pattern may be useful in distinguishing normal corneas from
keratoconic corneas.
Figure 1: Epithelial thickness maps and B-Scan for three subjects
(Normal Eye NC112, Early Keratoconus CP131, Severe Keratoconus
CP140)
Table 1
Commercial Relationships: Esther Young, Carl Zeiss Meditec,
Inc.; Homayoun Bagherinia, Carl Zeiss Meditec, Inc.; Patricia Sha,
Carl Zeiss Meditec, Inc. (C); Mary K. Durbin, Carl Zeiss Meditec,
Inc.
Non-invasive Assessments of the Sclera and Cornea after
Collagen Crosslinking and Glycosaminoglycan Depletion using
Multi-modal MRI
Kevin C. Chan1, 2, Leon C. Ho1, 4, Ian A. Sigal2, 3, Ning-Jiun Jan2, 3,
Xiaoling Yang1, 2, Yolandi van der Merwe1, 3, Gadi Wollstein2, 3,
Joel S. Schuman2, 3. 1NeuroImaging Laboratory, University of
Pittsburgh, Pittsburgh, PA; 2UPMC Eye Center, Eye and Ear Institute,
Ophthalmology and Visual Science Research Center, Department
of Ophthalmology, University of Pittsburgh, Pittsburgh, PA;
3
Department of Bioengineering, Swanson School of Engineering,
University of Pittsburgh, Pittsburgh, PA; 4Department of Electrical
and Electronic Engineering, University of Hong Kong, Hong Kong,
China.
Purpose: The microstructural organization and composition of
the corneoscleral shell are central to ocular biomechanics, and are
important in diseases such as glaucoma and myopia as well as in
corneoscleral treatments. However, limited non-invasive imaging
techniques are available to assess these properties globally and
quantitatively. Among the advanced magnetic resonance imaging
(MRI) techniques, diffusion tensor MRI (DTI) measures water
diffusion patterns and can reveal microstructural organization such as
primary fiber orientation, directionality and directional diffusivities,
whereas magnetization transfer MRI (MTI) is sensitive to the
concentration, mobility and surface chemistry of the macromolecules.
We hypothesized that multi-modal MRI can reveal the effects of
collagen crosslinking (mimicking aging conditions and corneoscleral
stiffening treatment) and glycosaminoglycan depletion (mimicking
glaucoma and myopic conditions) on the ocular tissues.
Methods: Sixteen freshly prepared ovine eyes were dissected to
give 4 sclera and 4 cornea tissue stripes (4-8 mm long, 1 mm wide)
per eye. The tissue stripes were treated with various concentrations
of glyceraldehyde solutions for collagen crosslinking (8 eyes) and
chondroitinase-ABC solutions for glycosaminoglycan depletion (8
eyes) at 37oC for 12 hours, followed by T2-weighted MRI, DTI and
MTI assessments using a 9.4-Tesla MRI scanner.
Results: Increasing concentrations of glyceraldehyde (Fig. 1) and
chondroitinase-ABC treatments (Fig. 2) decreased diffusivities and
increased magnetization transfer effect in the cornea. Glyceraldehyde
also increased magnetization transfer effect in the sclera (Fig.
1). No significant difference in fractional anisotropy by DTI was
detected in the cornea or sclera among different glyceraldehyde or
chondroitinase-ABC concentrations.
Conclusions: The increased magnetization transfer and reduced
diffusivities in the glyceraldehyde-treated ocular tissues suggested
that MTI and DTI could characterize the state of collagen
crosslinking in the eye under different physiological conditions.
Multi-modal MRI may also detect glycosaminoglycan removal
which is likely a cause of altered creep and stiffness in glaucoma
and myopic eyes. Multi-modal MRI can be useful for evaluating
the biomechanical and pathophysiological mechanisms in the
corneoscleral shell non-invasively and quantitatively.
Commercial Relationships: Kevin C. Chan, None; Leon C. Ho,
None; Ian A. Sigal, None; Ning-Jiun Jan, None; Xiaoling Yang,
None; Yolandi van der Merwe, None; Gadi Wollstein, None;
Joel S. Schuman, Zeiss, Inc. (P)
Support: National Institutes of Health Contracts P30-EY008098,
R01-EY023966 and UL1-TR000005 (Bethesda, Maryland);
BrightFocus Foundation G2013077 (Clarksburg, Maryland); Alcon
Research Institute Young Investigator Grant (Fort Worth, Texas);
Eye and Ear Foundation (Pittsburgh, Pennsylvania); and Research to
Prevent Blindness (New York, New York).
Corneal subbasal nerve plexus Decreases in Limbal stem cell
deficiency patients
Pichaya Chuephanich, Chantaka Supiyaphun, Carolina Aravena
Perez, Tahir Kansu Bozkurt, Fei Yu, Sophie X. Deng. Ophthalmology,
Jules Stein Eye Institute, Los Angeles, CA.
Purpose: To investigate the correlations between subbasal nerve
density (SND) in the cornea in limbal stem cell deficiency (LSCD)
patients by using in vivo laser scanning confocal microscopy.
Methods: Cross-sectional comparative study consisted of confocal
images of 51 eyes of 37 patients with LSCD collected by the
Heidelberg Retina Tomograph III Rostock Corneal Module Confocal
Microscope from 2009 to 2014 were analyzed. The scans of the
central cornea focused on the nerves at the subbasal epithelial layer
were evaluated by two independent observers. Seventeen normal eyes
of 13 patients served as control. Total SBN and long nerve density
(≥ 200 µm) were quantified. Statistical analysis was performed using
Wilcoxon rank-sum test and Kruskal-Wallis test.
Results: The total SND and long nerve density in the overall LSCD
groups were 50.2 ± 32.6 and 10.0 ± 10.7 nerves/cm2 which were
statistically reduced (P < 0.0001) compared with 97.3 ± 29.9 and
35.3 ± 25.3 nerves/cm2, respectively in the control group. The
percentages of SND reduction were 34.7% in early stage, 49.3% in
intermediate stage, and 78.3% in late stage of LSCD, respectively,
compared to the control group. The degrees of tortuosity of the nerve
differed significantly among the early, intermediate, and late stage of
LSCD patients. Total SND and long nerve density correlated with the
severity of LSCD.
Conclusions: Total SND and long nerve density could be used as one
of the quantifiable parameters to measure the degree of LSCD.
Commercial Relationships: Pichaya Chuephanich, None;
Chantaka Supiyaphun, None; Carolina Aravena Perez, None;
Tahir Kansu Bozkurt, None; Fei Yu, None; Sophie X. Deng
Pediatric Anterior Segment Ultrasound Biomicroscopy Image
Analysis: Intra and Inter-observer Agreement
Haoxing Chen1, Azam Qureshi1, Osamah Saeedi2, 3,
Mona A. Kaleem2, 3, Jordan Margo2, 3, Sachin Kalarn3,
Janet D. Leath2, 3. 1University of Maryland School of Medicine,
Baltimore, MD; 2Department of Ophthalmology, University
of Maryland Eye Associates, Baltimore, MD; 3Department of
Ophthalmology and Visual Sciences, University of Maryland School
of Medicine, Baltimore, MD.
Purpose: Previous studies have validated the use of ultrasound
biomicroscopy (UBM) to measure anterior segment (AS) structures
in adults. This is a novel study depicting a standardized imaging
and measurement protocol with reliability analysis using UBM and
ImageJ software.
Methods: Patients undergoing exam under anesthesia for AS
pathology (including cataract, glaucoma, and dysgenesis) were
imaged using UBM. Using the trabecular-iris angle (TIA) as a
landmark reference point, 40 UBM images from 20 pediatric patients’
eyes were analyzed by 4 trained observers. 46 structural parameters,
some of which relied on the TIA as a landmark and others did not
(TIA-dependent [TD] and non-TIA dependent [NTD], respectively),
were measured in each image twice using ImageJ by the same
observer. Some measurements required reference lines/measurements
to be taken first and were classified as a certain degree removed from
the reference (seen in Tables 1 & 2). Corresponding measurements
were compared between observers. Intra-observer repeatability
(IOR) and inter-observer agreement (IOA) for the measurements
were determined using coefficient of variation (CV) and intra-class
correlation (ICC) followed by assessment of Bland-Altman plots
(BAP) for each pair of observers, respectively.
Results: For NTD parameters, non-ciliary body (CB) related
measurements showed CV range 0.60-16.22% and ICC range 0.840.89, whereas CB-related parameters showed CV range 2.86-23.40%
and ICC range 0.29-0.92. For TD parameters, parameters < 2 degrees
removed from reference showed CV range 0.024-5.40% and ICC
range 0.89-1.00, whereas parameters > 1 degree removed showed
CV range 0.63-27.44% and ICC range 0.22-1.00. Results shown in
Tables 1 & 2. BAP’s demonstrated no variation in reliability related
specifically to measurements’ underlying values.
Conclusions: Preplaced landmarks yielded good IOR and IOA in
quantitative assessment of AS structures that were NTD and non-CB
related, or less removed from reference. NTD measurements based on
CB varied greatly in IOR and IOA, suggesting protocol modifications
to improve accuracy or qualitative assessment for the CB. Increasing
variability in TD measurements with increasing degrees removed
from the reference landmark supports use of another reference
landmark to minimize variation. This study provides an assessment
of reliability of AS measurements that may be applied to UBM image
analysis.
Commercial Relationships: Haoxing Chen, None; Azam Qureshi,
None; Osamah Saeedi, None; Mona A. Kaleem, None;
Jordan Margo, None; Sachin Kalarn, None; Janet D. Leath, None
Support: Knights Templar Eye Foundation Career Starter Grants
2015-2016: Anterior Segment Ultrasound Biomicroscopy Findings
and Implications in Children with Congenital, Aphakic, and
Pseudophakic Glaucoma
Assessment and Comparison of Automated Anterior Chamber
Metrics between Casia Swept-Source OCT vs Pentacam
Scheimpflug Imaging
Kenneth Marion1, 2, Anna Dastiridou1, 2, Brian Francis1, 2,
Vikas Chopra1, 2. 1Doheny Eye Institute, Los Angeles, CA; 2University
of California Los Angeles, Los Angeles, CA.
Purpose: To better understand the automated tools provided by two
anterior segment imaging technologies and to determine if light has
an effect on Anterior Chamber (AC) Volume
Methods: 26 normal and 26 POAG participants (n=103 eyes) had
both eyes imaged twice using the Tomey Casia SS1000 SweptSource OCT ‘3D Anterior Segment’ scan (Aichi, Japan) and
Oculus Pentacam ‘3D Scan HR’ scan (Arlington, WA) at low light
conditions (<0.2fce) at Doheny Eye Centers-UCLA (Figure 1).
Gonioscopy showed open angles (grade III-IV on Schaffer scale) for
all participants. 29 participants (n=46 eyes) were additionally imaged
at normal room light conditions (8.0fce) with the Casia. Sper Light
Meter FC–840021 (Scottsdale, AZ) was used to measure light levels
at the eye/camera interface.
The software tools auto-segment ocular layers and calculate AC
Volume (ACV) and Pupil Diameter (PD). Participants were rescanned
if mistracing issues were present.
Statistics for instrument agreement and correlation (averaged the two
acquisitions for each device) and low/normal light metrics using the
Casia were determined with correlation coefficients and paired t-tests
using MedCalc v13.1.2.0.
Results: The mean±SD Casia ACV/PD was
153.7±45mm3/4.7±1.2mm and Pentacam
148.5±47.2mm3/2.8±0.9mm. ACV/PD correlation between Casia
and Pentacam was 0.91/0.56, respectively. There was a statistically
significant difference between the ACV for the devices (p=0.01).
The mean±SD Casia ACV/PD with the Lights ON in the imaging
room was 159.3±36.9mm3/5.9±1.1mm and Lights OFF was
154.7±35.1mm3/3.97±0.81mm. ACV/PD correlation between
Casia Lights ON and OFF was 0.92/0.74, respectively. There was
a statistically significant difference for the ACV at varied lighting
conditions (p=0.04).
Conclusions: Reliable AC volumes can be assessed using both Casia
OCT and Pentacam as determined by automated AC metrics analyses.
Although statistically significant differences for ACV are found
between devices, the differences are small with good correlation with
matched lighting conditions. Changes in lighting can affect the ACV
& PD values and thus strict lighting conditions should be employed
to achieve reliable longitudinal comparisons. Our study underscores
the notion that adoption of devices for clinical evaluation for AC
parameters will only occur if the devices provide reliably automated
analyses.
Figure 1: A) Casia B) Pentacam Volume Scans
Commercial Relationships: Kenneth Marion, None;
Anna Dastiridou, None; Brian Francis, Innfocus (F), BVI
Endooptiks (C), Lumenis (F), Aquesys (F), Allergan (F);
Vikas Chopra, Allergan (F)
Anterior chamber angle imaging with Automatic
Gonio -Photography
Carlo Alberto Cutolo1, Alessandro Bagnis1, Riccardo Scotto1,
Chiara Bonzano1, Pietro Frascio1, Michele Pascolini2,
Cesare Tanassi2, Carlo E. Traverso1. 1University Eye Clinic,
University of Genoa, Genoa, Italy; 2NIDEK Technologies Srl, Padua,
Italy.
Purpose: Assessing the anterior chamber angle (ACA) is essential
when evaluating patients with suspected or diagnosed glaucoma.
The aim of this study was both to assess inter-observer agreement on
Automatic Gonio-Photography (AGP) - true color images of the ACA
- and inter-device agreement by comparing AGP to gonioscopy.
Methods: A prototype AGP device (Gonioscope NGS-1; Nidek
Technologies srl) was used to acquire 360-degree images on both
eyes of twenty consecutive patients recruited from the glaucoma
service at the University Eye Clinic of Genoa (Italy). Two masked
observers graded the apparent iris insertion for each of the four
quadrants of the ACA and reported ACA abnormalities in AGPs
randomly presented. All patients underwent dynamic gonioscopy and
the 4 quadrants were graded again using the Spaeth Classification.
Inter-observer and inter-device agreement for apparent iris insertion
were determined by using Cohen’s linearly weighted κ (KW)
coefficient of concordance. Statistical analysis was performed using
MedCalc 15.11 (MedCalc Software, Ostend, Belgium).
Results: Twenty (12.5%) of the 160 quadrants were excluded
from statistical analysis because of poor image quality. AGP
showed substantial inter-observer agreement (KW, 0.77; 95% CI,
0.67-0.87) with regards to apparent iris insertion. Both observers
correctly identified ACA abnormalities, i.e. iridotrabecular contact
in two or more quadrants (5), iridectomy (3), internal ostium of the
trabeculectomy site (2), EX–PRESS® device (1), tube (1), anterior
chamber IOL (1), and angle recession (1). All abnormalities were
confirmed at gonioscopy. Differentiating between appositional and
synechial iridotrabecular contact was only possible with dynamic
gonioscopy. Results of AGP and gonioscopy showed almost perfect
inter-device agreement on apparent iris insertion evaluation (KW,
0.92; 95% CI, 0.86 to 0.98).
Conclusions: AGP using the NGS-1 gonioscope is a reliable method
for assessing apparent iris insertion and proved useful in detecting
pathological and postoperative ACA findings in glaucoma patients.
This technique also appears to be reliable when recording ACA
structures.
Angle recession obtained with AGP.
Commercial Relationships: Carlo Alberto Cutolo, None;
Alessandro Bagnis; Riccardo Scotto, None; Chiara Bonzano,
None; Pietro Frascio, None; Michele Pascolini, NIDEK
Technologies Srl; Cesare Tanassi, NIDEK Technologies Srl;
Carlo E. Traverso, None
Imaging of aqueous outflow system structures using
ultrahigh -resolution optical coherence tomography
Rene M. Werkmeister1, Sabina Sapeta1, Valentin Aranha dos Santos1,
Gerhard Garhofer2, Leopold Schmetterer1, 2. 1Center for Medical
Physics and Biomedical Engineering, Medical University of Vienna,
Vienna, Austria; 2Department of Clinical Pharmacology, Medical
University of Vienna, Vienna, Austria.
Purpose: To investigate the capability of ultrahigh-resolution OCT
for visualizing the structures of the aqueous outflow system (AOS),
i.e. Schlemm’s canal and collector channels and to evaluate changes
in the lumen area of the outflow system after topical instillation of
pilocarpine.
Methods: An ultrahigh-resolution OCT system, comprising a broad
band Ti:Sapphire laser operating at 800 nm (spectral bandwidth 170
nm) and a high speed CMOS camera with a maximum read out rate
of 140 kHz for imaging of the anterior chamber was employed in the
present study. The system offers an axial resolution in tissue of about
1.2 μm (lateral resolution approx. 16 μm), allowing for visualization
of details of the corneal anatomy with high precision. In 15 healthy
subjects, three dimensional data sets covering 4 x 4 x 2 mm3 and
512 x 512 x 2048 pixels were recorded in different quadrants of the
limbal area in order to image AOS structures. Pilocarpine was used
as a model drug to induce an increase in the lumen diameter of these
structures. In the obtained data sets, segmentation of Schlemm’s canal
was performed in order to evaluate the lumen area before and after
instillation of one drop pilocarpine.
Results: Schlemm’s canal and collector channels could be visualized
and segmented using ultrahigh-resolution OCT and custom software
written in Matlab. Average lumen area of Schlemm’s canal was
estimated to be 7500 +/- 2540 µm2 (Figure 1(a)). After topical
instillation of one drop pilocarpine, the average lumen area increased
to 12200 +/- 3490 µm2 (Figure 1(b)).
Variability of the lumen area estimation was calculated to be 5 %.
Conclusions: Ultrahigh-resolution OCT is well suited to study the
aqueous outflow system and to evaluate the lumen area of Schlemm’s
canal. In glaucoma, the architecture of the AOS and its diameter
could be an important factor for the future classification, therapy,
evaluation of the progression risk and prognosis. Based on the
obtained measurement protocol, further studies investigating anatomy
and pathophysiology of the anterior segment of the eye as well as
surgical outcome in patients with glaucoma can be planned.
Fig. 1: OCT images of the limbal region of a healthy volunteer (a)
before and (b) after topical instillation of one drop pilocarpine.
Commercial Relationships: Rene M. Werkmeister,
None; Sabina Sapeta, None; Valentin Aranha dos Santos;
Gerhard Garhofer, None; Leopold Schmetterer, None
Support: The Heidelberg Engineering Academy Glaucoma Research
Initiative
Blood-neural-barrier disruption has different effects on
fluorescein angiography dynamics in the eye and brain
Flora Hui1, Christine T. Nguyen1, Zheng He1, Rachel Gurrell2,
Rebecca Fish2, Algis J. Vingrys1, Bang V. Bui1. 1University of
Melbourne, Melbourne, VIC, Australia; 2Neuroscience and Pain
Research Unit, Pfizer, Cambridge, United Kingdom.
Purpose: The blood-retinal-barrier may be a surrogate for assessing
blood-brain-barrier integrity. If so, systemic disruption of tight
junctions should produce similar effects on fluorescein angiography
dynamics in both tissues. We compare the angiography dynamics
in rat retina and brain before and after blood-neural-barrier (BNB)
disruption.
Methods: Sodium fluorescein (1%, 70μl, 1.05ml/min) was injected
via the femoral vein in anesthetized (60:5mg/kg ketamine:xylazine)
adult Long-Evans rats and angiography recorded (Micron III) at
30 frames/sec for 1 min. An area of skull was thinned to image
vasculature on the cortical surface. Angiography of naïve eyes (n=26)
and brains (n=18) established baseline profiles.
Angiography was performed at 6 and 24 hours after BNB disruption
with IV sodium deoxycholate (1ml 0.06M DOC, 33μl/min, n=10/
group). Controls received saline (1ml). Each pixel’s time-luminance
profiles were analysed (MATLAB) to return time to 50% maximum
(filling), 50% decay (drainage) and fluorescence at 1 min (residual,
%). Frequency histograms described the distribution of indices
before and after BNB disruption. Level of injury was quantified by
the percentage of pixels showing indices beyond the upper quartile
(>75%) of control.
Results: Angiography luminance profiles have similar shapes in both
tissues, with a rapid increase in fluorescence and exponential-like
decay. Cortical arteries filled faster (all p<0.05) than retinal arteries
(mean±SEM, brain 5.6±0.2s, eye 6.6±0.1s) and also decayed faster
(brain 10.3±0.2s, eye 11.0±0.1s). The residual was significantly
brighter in the brain (brain artery 34.7±2.7%, eye 12.3±0.3%).
Similar differences were observed in capillaries and veins (p<0.05).
Following DOC injection fluorescein filling and decay was similar at
6 and 24 hours in both tissues. The residual in capillaries, an indicator
of vascular leakage, was elevated compared to control at 6 hours in
the eye (brain -0.1±8.9%, eye +22.3±10.6%) but at 24 hours in the
brain (brain +40.5±10.1%, eye -18.2±2.9%).
Conclusions: Spatial and temporal analysis of fluorescein
angiography in the eye and brain showed subtle differences in filling
and decay. Cortical vessels were brighter due to lack of pigment
in the brain. Disruption of tight junctions with DOC affected the
BNB of the eye earlier than the brain (6 vs 24 hours) possibly due to
differences in blood flow and BNBs in the two tissues.
Commercial Relationships: Flora Hui, None;
Christine T. Nguyen, None; Zheng He, None; Rachel Gurrell;
Rebecca Fish, Pfizer; Algis J. Vingrys, None; Bang V. Bui, None
Support: Australia Research Council Linkage Project Grant
100200129, Australian Postgraduate Award (Industry)
Modeling and experimental testing of iris thermal damage limits
David Dewey, Alexander Vankov, Georg Schuele. R&D, Abbott
Medical Optics, Sunnyvale, CA.
Purpose: Despite that retinal safety limits are well established and
useful tool for scanning laser systems, iris thermal damage threshold
is not well described and understood. Here we present experimental
results and theoretical data for iris damage threshold for laser system.
Methods: Experimental evaluation of iris minimal visible lesions
was done on fresh (less than 25h post mortem) porcine eyes with
removed cornea placed under water immersion. Laser beam with
180um in diameter and variable spot spacing was scanned over the
iris surface, creating rectangular lesions of various sizes. These
lesions were analyzed under the surgical microscope to determine the
minimal visible lesion threshold, see Fig.1.
Analytical Point Spread Function solution and Finite Element Mesh
(FEM) methods were independently used for temperature calculation.
Temperature rise was calculated for 1uJ 70kHz laser beam exercising
lawn-mower pattern with variable spot spacing 1-3um and distant
boundary conditions at body temperature. The analytical solution was
used to verify the FEM method and all results were compared the
experimental MVL threshold data.
Results: We found that thermal steady state is achieved in fraction
of a second. Maximum temperature is reached in the center of the
exposed area. Temperature profile consists of slow temperature
change and fast oscillations, arising when beam is passing through
the point of interest (Fig.2). At the same laser power, maximal
temperature slowly depend on the spot spacing and length of the cut,
but inversely proportional to the width.
Conclusions: Experimental results matches theoretical model for
minimal visible lesion prediction. Based on these results, one can
calculate parameters of laser scanning system for various incision
geometries.
© 2015 Abbott Medical Optics Inc.
Figure 1: Thermal lesions created on the iris at different laser powers.
Figure 2: Calculated FEM temperature profiles of the points at the
center line of the exposed area at different distance to the center.
Commercial Relationships: David Dewey, Abbott Medical Optics
(P), Abbott Medical Optics; Alexander Vankov, Abbott Medical
Optics (P), Abbott Medical Optics; Georg Schuele, Abbott Medical
Optics (P), Abbott Medical Optics
Measurement of Iris Lesion Volume with Anterior Segment
Swept-Source Optical Coherence Tomography
Laura Vickers1, Handan AKIL1, 2, Elmira Baghdasaryan1, 2,
Olivia L. Lee1, 2. 1Ophthalmology, Doheny Eye Institute-UCLA,
Pasadena, CA; 2Doheny Image Reading Center, Los Angeles, CA.
Purpose: To calculate the volume of iris lesions using anterior
segment swept-source optical coherence tomography (OCT) to
enhance the clinical assessment of lesions, with comparison to the
gold standard ultrasound biomicroscopy (UBM).
Methods: Two eyes of two patients with iris lesions underwent
imaging with anterior segment swept-source OCT (Casia SS-1000,
Tomey, Nagoya, Japan) as well as ultrasound biomicroscopy (Ellex
Eyecubed 40 MHz UBM, Ellex medical Pty. Adelaide, Australia).
Image processing software ImageJ (U. S. National Institutes of
Health, Bethesda, Maryland, USA) was used to calculate the
iris lesion volume from the 3-dimensional (3D) gonioscopic
reconstruction of the OCT images. Lesion volume as calculated from
the OCT was compared to lesion height as measured by UBM.
Results: One iris lesion, a clinically diagnosed iris melanoma,
measured 5.7 mm in greatest circumference and 5.5 µL volume. The
second iris lesion, clinically diagnosed as a large cyst, involved both
the ciliary body and iris, measuring 15.4 mm greatest circumference
and 76.0 µL volume.
Conclusions: Anterior segment OCT with 3D gonioscopic
reconstruction can be used to calculate the volume of iris lesions.
This may be used for clinical monitoring of lesions including change
in size over time. Future studies of a greater number of eyes over
time will compare sensitivity of OCT versus UBM for detection of
change in size of iris lesions.
Commercial Relationships: Laura Vickers, None; Handan AKIL;
Elmira Baghdasaryan, None; Olivia L. Lee, None
Video imaging during visual field perimetry
Jacques R. Charlier1, Xavier Zanlonghi2. 1Research, Metrovision,
Perenchies, France; 2Clinique Sourdille, Nantes, France.
Purpose: Video imaging consists in recording the entire visual field
process in synchrony with the video of the patient’s head. Several
clinical applications have been investigated to evaluate the clinical
usefulness of this new technology.
Methods: This study included results from 48 visual field exams
performed on a MonCvONE full field projection perimeter with
synchronized video recording.
The video from a large viewing field camera was recorded in
synchrony with the position of the visual stimulus, with other
test parameters such as luminance and size and with the patient’s
response obtained from the patient’s press button or from the operator
judgment. The study included patients who were unable to perform
automated perimetry due to young age or handicap, patients with
abnormal eye movements, head posture or ptosis and controls
performed after automated perimetry.
Results: Video recording was extremely useful in the majority of
clinical cases. 24 exams were performed on young children (age
between 2 and 5 years) using attraction perimetry. The eye orientation
responses could be interpreted and validated after the exam. In
other cases, the video recording facilitated the interpretation and
documentation of visual field results with the inclusion of video
snapshots in the examination report. Additional applications included
the recording of cardinal eye gaze positions and of the fusion visual
field.
Conclusions: Synchronized video imaging performed during visual
field exams is a clinically useful tool for the examination of patients
who cannot perform automated perimetry and for the documentation
of artefacts and situations such as ptosis, abnormal eye movements,
abnormal head posture and incorrect position of refraction correction.
Commercial Relationships: Jacques R. Charlier, Metrovision (I);
Xavier Zanlonghi, None
Support: Eurostars grant 6034
Ocular biometric parameters variation between closed and
opened angles in Thai population
Kulawan Rojananuangnit, Panidaporn Salyapongse. Ophthalmology,
Mettapracharak(Wat Rai Khing) hospital, Nakhon Pathom, Thailand.
Purpose: To compare the ocular biometric parameters between
closed angle and opened angle by using anterior segment OCT
Methods: This was a cross sectional descriptive study. We enrolled
normal, glaucoma suspected and glaucoma participants with visual
acuity better than 20/70, refraction ± 6 diopters and astigmatism ± 3
diopters and excluded secondary causes of glaucoma. All consented
participants were completely performed eye examination and
measured the following ocular biometric parameters including, axial
length, central corneal thickness, gonioscopy and Visante®anterior
segment optical coherence tomography. Linear regression model was
used to test the differences between groups.
Results: Two-hundred twenty three opened angle eyes (112
participants) and 50 closed angle eyes (25 participants) were
recruited. The mean age of closed group was significant older than
opened group, 63.58 ± 9.06 versus 56.75 ± 13.39 years (p=0.022).
Refraction between 2 groups were similar; however, the axial length
were significant longer in the opened group; 23.49 ± 0.93 mm
versus 22.56 ± 0.92 mm (p<0.05). Intraocular pressure and central
corneal thickness were eqivalence. All anterior segment parameters
in both nasal and temporal sided of opened group were higher
including; anterior chamber depth (ACD) 2.92 ± 0.56mm versus
2.41 ± 0.73mm, p<0.001; temporal angle opening distance 500 μm
(AOD500) 0.78 ± 0.33 mm versus 0.5 ± 0.19 mm, p<0.001; nasal
AOD500 0.64 ± 0.23mm versus 0.42 ± 0.15 mm, p<0.001; temporal
AOD750 0.97 ± 0.36 mm versus 0.67 ± 0.25mm, p<0.001; nasal
AOD750 0.81 ± 0.25 mm versus 0.54 ± 0.16mm, p<0.001; temporal
trabecular–iris space area 500 μm(TISA500) 0.31 ± 0.15mm2 versus
0.19 ± 0.08mm2, p<0.001; nasal TISA500 0.24 ± 0.1mm2 versus 0.16
± 0.06mm2, p<0.001; temporal TISA750 0.53 ± 0.23mm2 versus0.34
± 0.13mm2, p<0.001; nasal TISA750 0.42 ± 0.15mm2 versus 0.27 ±
0.09mm2, p<0.001; temporal scleral spur angle 54.39 ± 10.52 degree
versus 43.35 ± 10.97 degree, p<0.001; nasal scleral spur angle 49.34
± 10.25 degree versus 38.06 ± 9.4 degree, p<0.001.
Conclusions: Our study provided anterior segment parameters of
opened and closed angle in Thai population which can consider as the
normative reference.
Commercial Relationships: Kulawan Rojananuangnit, None;
Panidaporn Salyapongse
Novel fshape mapping of peripapillary RNFL and choroidal
thickness for group-wise comparison of glaucoma and healthy
aging
Sieun Lee1, Karteek Popuri1, Joanne A. Matsubara2, Brennan Eadie2,
Andrew Merkur2, Paul MacKenzie2, Marinko V. Sarunic1, Mirza
Faisal Beg1. 1Simon Fraser University, Vancouver, BC, Canada;
2
University of British Columbia, Vancouver, BC, Canada.
Purpose: The functional shape (fshape) registration framework maps
the retinal layer thickness measurements from multiple subjects
into a common mean template space, allowing for a cross-sectional
analysis in greater spatial detail than conventional point-wise or
sectoral comparison. Using fshape mapping, we examined the group
differences in the RNFL and choroidal thickness by age and presence
of glaucoma.
Methods: 10 young healthy eyes (Group A, mean age: 29.8 ±
3.6), 10 older healthy eyes (Group B, mean age: 57 ± 4.4), and 18
older glaucomatous eyes (Group C, mean age: 61.7 ± 7.9) were
imaged with a prototype swept-source OCT. The retinal nerve fiber
layer (RNFL) and choroid were automatically segmented and their
thickness was computed. Given a group of retinal surfaces, the fshape
framework simultaneously estimates the mean template surface and
mean thickness measurement mapping on the template. Using this
framework, mean thickness maps were generated for each group,
followed by point-wise Welch’s t-tests to identify the regions of
significant difference due to healthy aging or glaucoma.
Results: Figure 1-a shows the group-mean RNFL thickness maps.
All groups display the hourglass pattern with overall thinner RNFL
in Group C. Figure 1-b shows group difference maps by age (left)
and presence of glaucoma (right), with the latter displaying a pattern
similar to the healthy RNFL thickness maps in Figure 1-a. Figure
1-c shows the p-values of the difference maps in Figure 1-b. The
glaucoma-based difference (right) shows the highest significance in
the inferior region.
Figure 2-a shows the group-mean choroidal thickness maps. In all
groups, the choroid is thinner in the inferior region, and thicker in
the nasal and superior regions. Group B shows the overall thinner
choroid than Group A (Figure 2-b & c, left), with the nasal region
showing the largest difference and statistical significance. Group
C shows the thinner choroid than Group B (Figure 2-b & c, right),
although the effect of glaucoma appears to be less than that of age. In
both layers, the differences are the most pronounced where the layers
are originally the thickest.
Conclusions: We performed group-wise comparison of the RNFL
and choroidal layers in healthy aging and glaucoma using the novel
fshape registration framework, showing detailed regional patterns not
detectable by conventional point-wise or sectoral comparison.
Imaging of scar tissue in the filtering bleb by anterior-segment
polarization-sensitive OCT
Masahiro Yamanari1, Satoru Tsuda2, Taiki Kokubun2, 3,
Kazuko Omodaka2, Yu Yokoyama2, Noriko Himori2, Shiho KunimatsuSanuki2, Kazuichi Maruyama2, Hiroshi Kunikata2, Toru Nakazawa2.
1
Tomey Corporation, Nagoya, Japan; 2Department of Ophthalmology,
Tohoku University Graduate School of Medicine, Sendai, Japan;
3
Department of Ophthalmology, Katta General Hospital, Shiroishi,
Japan.
Purpose: Wound healing and scarring in filtering bleb created by
trabeculectomy have important roles in decreased functioning of the
filtering bleb, although these changes cannot be detected in standard
imaging methods. The purpose of this study is to demonstrate
anterior-segment polarization-sensitive optical coherence tomography
(AS-PS-OCT) for visualization of the filtering bleb and healthy
anterior eye segment.
Methods: A prototype AS-PS-OCT with a swept source at 1.3 µm
wavelength was developed. Anterior segment around the angle was
measured in four healthy human eyes, and filtering blebs of two
glaucomatous human eyes that had nonfunctioning filtering bleb
were measured. PS-OCT data were processed using Cloude-Pottier
decomposition to estimate unbiased local phase retardation of
birefringent tissue and to calculate entropy that shows randomness
of polarization property. In addition, attenuation coefficient image
was calculated and used to create composite images with the phase
retardation or entropy images.
Results: In all healthy eyes, conjunctiva and sclera showed weak and
moderate phase retardations, respectively (Fig. 1(c)). The composite
image in Fig. 1(d) was effective to visualize both light scattering
and birefringence of the tissues in Fig. 1(b) and (c), respectively.
Entropy images in Fig. 1(e) and (f) showed low, moderate and high
randomness of polarization properties in conjunctiva, sclera and
uvea, respectively. In all glaucomatous eyes, regions with high phase
retardation were found in the bleb wall, indicating scar tissue with
highly organized collagen fibers. In Fig. 2(c) and (d), not only the
large scar tissue at the adhesion between the bleb wall and scleral flap
but also fine and diffuse scar tissues were found in the bleb wall.
Conclusions: AS-PS-OCT with the advanced signal processing
provides additional contrast in both healthy tissue and scarred
filtering bleb. The composite images using the attenuation coefficient
enabled comprehensive understanding of the tissue properties in light
scattering and birefringence.
Commercial Relationships: Sieun Lee; Karteek Popuri,
None; Joanne A. Matsubara, None; Brennan Eadie,
None; Andrew Merkur, None; Paul MacKenzie, None;
Marinko V. Sarunic, None; Mirza Faisal Beg, None
Support: Brain Canada; Natural Sciences and Engineering Research
Council of Canada (NSERC); Canadian Institutes of Health Research
(CIHR); Michael Smith Foundation for Health Research (MSFHR);
Alzheimer Society Canada; Pacific Alzheimer Research Foundation;
Genome British Columbia
Figure 1: Images of the healthy anterior eye segment. OCT intensity
(a), attenuation coefficient (b), local phase retardation (c), composite
image (d) of (b) and (c), entropy (e), composite image (f) of (b) and
(e) are shown.
Figure 2: Images of the filtering bleb. The notation is same as Fig. 1.
Commercial Relationships: Masahiro Yamanari, Tomey
Corporation (P), Tomey Corporation; Satoru Tsuda, None;
Taiki Kokubun, None; Kazuko Omodaka, None; Yu Yokoyama,
None; Noriko Himori, None; Shiho Kunimatsu-Sanuki,
None; Kazuichi Maruyama, None; Hiroshi Kunikata, None;
Toru Nakazawa, Tomey Corporation (F)
Assessment of diagnostic information acquired by Transverse
Section Enhanced Depth Imaging Optical Coherence
Tomography in Patients with Optic Nerve Head Drusen
Katharina Blobner, Mathias M. Maier, Nikolaus Feucht. Department
of Ophthalmology, Technische Universität München, Munich,
Germany.
Purpose: Optic head nerve drusen (ONHD) are less frequently
detected in clinical routine, than they actually exist. The pathology
of ONHD is a proceeding calcification in the optic nerve.
Conventionally ophthalmoscopy, fundus autofluorescence (FAF)
and B scan ultrasound are used to identify and illustrate ONHD.
Transverse section enhanced depth imaging optical coherence
tomography (TSV-EDI-OCT) enables imaging of deeper layers of
the subretinal structures. The purpose of this study was to describe
the morphologic information acquired by TSV-EDI-OCT, to compare
the diagnostic information of this method to previous imaging
techniques. Furthermore TSV-EDI-OCT was used to detect ONHD
and acquire measurements regarding to the surrounding anatomical
constitutions.
Methods: 23 eyes of 12 patients (6 women, 6 men, mean age
64 years) with diagnosed ONHD were included. A complete
ophthalmological examination, perimetry, B scan ultrasound,
FAF and TSV-EDI-OCT through the optic nerve were performed.
Localizations of drusen in regard to Bruchs Membrane, greatest
linear diameter in Bruchs Membrane Opening (BMO), choroidial
thickness in different eccentricities and papilla prominence were
measured.
Results: All eyes presented ONHD in ultrasound examination. TSVEDI-OCT identified ONHD in 22 eyes, 21 eyes with superficial and
19 eyes with buried drusen. In these 22 eyes hyperreflective plaques
were demonstrated. Not only the plaques but also the different
reflectivity of the drusen volume enables a clear delimitation to
other structures primarily to vessels. The mean horizontal diameter
of BMO was 1594 (± 189) µm. 1500 µm nasal from the central of
the optic disc the choroidial thickness showed an average of 107
(± 28) µm and temporal 110 (± 27) µm. A mean maximal papilla
prominence of 661 (± 149) µm was found. Importantly, TSV-EDIOCT was able to verify buried drusen in 5 eyes, which did not show
any corresponding signal in FAF.
Conclusions: B scan ultrasound is still gold standard in patients
with ONHD. But particularly regarding to the anatomically position
and surrounding area TVS-EDI-OCT provides valuable detectability
and monitoring of ONHD. In addition, TVS-EDI-OCT illustrates
hyperreflective plaques and may permit a more sensitive proof of
buried drusen than FAF.
Commercial Relationships: Katharina Blobner, None;
Mathias M. Maier, None; Nikolaus Feucht
Three Dimension Spectral Domain Optical Coherence
Tomography: an Important Tool for Diagnosis, Treatment and
Follow-up of Optic Disc Pit in Elderly Patients
Luis Alberto Zeman Bardeci, Matias Iglicki, Mariano Cotic,
Marcos Mendaro, Lucas Adamo, Jorge Bar, Pablo Chiaradia,
Marcelo Zas. División Oftalmología, Hospital de Clínicas, City of
Buenos Aires, Argentina.
Purpose: Our purpose was to study the clinical manifestations and
course of optic pit maculopathy using Spectral Domain Optical
Coherence Tomography (SD- OCT) in elderly patients who were
misdiagnosed for this condition.
The origin of the fluid and precise pathophysiology of optic disc pit
maculopathy remains unclear. It has been suggested that submacular
fluid originates either from vitreous or cerebrospinal fluid. We report
a case series of optic disc pit maculopathy which were unsuccessfully
diagnosed without using SD OCT. Using this technique we achieved
to distinguish this pathology, treat it and follow its evolution.
Methods: We examined 35 eyes of 32 patients with macular
detachment combined with optic disc pit. All patients were 65 years
old or older.
They were mistreated as neovascular membrane due to age-related
maculopathy.
The diagnosis and monitoring of patients were done by means of:
symptoms, Fluorescein Angiography, SD-OCT and Best Corrected
Visual Acuity (BCVA) with ETDRS charts. Patients visual acuity
prior to surgery was in average 20/300.
Results: We observed six different foveal appearances in regard to
fluid localization. We could clearly distinguish optic disc pit and
neovascular membrane using this technology. Fluid accumulation
was observed below the margin of the optic disc and hyperreflective
porous tissue was onserved in the optic disc excavation. We were able
to diagnose optic disc pit faster and better.
Conclusions: Considering that this OCT technique has no relevant
side effects, we strongly recommend using it in every patient who
has any kind of maculopahty, since optic disc pit could be a probable
differential diagnosis.
3-dimensional SD-OCT scans revealed a connection between
subretinal space, intraretinal space, perineural space, and the
vitreous cavity. Therefore, we suggest that intraretinal or subretinal
fluid in optic disc pit maculopathy may have both a vitreous and
cerebrospinal origin, as it has been published.
Distinguishing Ischemic Optic Neuropathy from Optic Neuritis
by Ganglion Cell Analysis
Carlos E. Mendoza1, 2, Natalie Erlich-Malona1, Nimesh Patel1, 3,
Caitlin Monaco1, Emily Cole1, Thomas Hedges1. 1New England
Eye Center, Tufts Medical Center. Tufts University, Boston, MA;
2
Neurology, New York University, New York, NY; 3Ophthalmology,
Bascom Palmer Eye Institute, Miami, FL.
Purpose: Optic neuritis (ON) and nonarteritic ischemic optic
neuropathy (NAION), two of the most common optic neuropathies
in adults, are often difficult to distinguish from one another at initial
presentation. Our aim was to determine whether an altitudinal pattern
of ganglion cell loss, as measured by Optical Coherence Tomography
(OCT), can be used to distinguish NAION from ON during the acute
phase, and whether the speed or severity of ganglion cell loss differs
between the two diseases.
Methods: We performed a retrospective, case-control study of
44 patients (50 eyes) with ON or NAION and 44 age-matched
controls (50 eyes). All subjects had retinal imaging with HD-OCT.
NAION and ON patients had OCT imaging at presentation and four
consecutive follow-up visits. Controls had OCT imaging at one point
in time. Mean ganglion cell complex (GCC) thickness, mean retinal
nerve fiber layer (RNFL) thickness, and GCC altitudinal difference
were compared between NAION and ON patients at each time point
and between disease and control subjects using unpaired t-tests.
Results: Mean time from onset to presentation was 11 days for
both NAION and ON. There was a significantly greater altitudinal
difference in GCC thickness in NAION patients than ON patients
at all time points (p=0.01 – 0.049). Mean GCC thickness was
decreased at 0-2 weeks in both NAION and ON compared to controls
(p<0.001). Mean RNFL thickness was increased in NAION at 0-2
weeks compared to controls (p=0.038) and did not significantly
differ between ON and controls at 0-2 weeks. GCC thickness was
significantly decreased in NAION compared to ON only at 16+
weeks (p=0.036). RNFL thickness was significantly increased in
NAION compared to ON at 3-4 weeks, but did not differ significantly
at any other time point.
Conclusions: Altitudinal difference in GCC thickness, as seen with
OCT, is significantly greater in NAION than in ON, representing a
key factor in distinguishing one disease from the other. Altitudinal
loss of ganglion cells supports a diagnosis of NAION, whereas ON
is characterized by more diffuse loss. Significant GCC thinning
can be observed in both ON and NAION within two weeks of
onset, confirming evidence that cell death begins early and must be
addressed as quickly as possible.
Commercial Relationships: Luis Alberto Zeman Bardeci, None;
Matias Iglicki, None; Mariano Cotic, None; Marcos Mendaro;
Lucas Adamo, None; Jorge Bar, None; Pablo Chiaradia, None;
Marcelo Zas, None
Results: Initial programming focused on specific features such as
disc detection, cup to disc ratio, probability of glaucoma, vessel
detection, and retinal thickness. These are then integrated at a
“higher level” of decision making to assist in the diagnosis. Several
conditions have been modeled and underwent preliminary testing.
Conclusions: The Queryable Atlas of the Retina is developing
into a universal retina database for ophthalmologists, trainees, and
photographers, allowing sharing, learning and collaboration and as a
prototype of diagnostic support system.
Commercial Relationships: Jessica Taibl, None; Samir I. Sayegh,
None
Altitudinal Difference in GCC Thickness.
A-C: OCT analysis at 4 weeks of a patient with NAION. D-F: OCT
analysis at 3 weeks of a patient with ON.
Commercial Relationships: Carlos E. Mendoza, None;
Natalie Erlich-Malona, None; Nimesh Patel, None;
Caitlin Monaco, None; Emily Cole, None; Thomas Hedges, None
Support: Research to Prevent Blindness Challenge grant to the New
England Eye Center/Department of Ophthalmology, Tufts University
School of Medicine
RetinAsk Update: Image and Image Analysis, Worth a Million
Words?
Jessica Taibl, Samir I. Sayegh. The EYE Center, Champaign, IL.
Purpose: To present the evolution of RetinAsk: The Queryable Atlas
of the Retina
Methods: Originally presented in 2000 (Sayegh et al, AAO 2000)
RetinAsk is unique as it is a queryable retina atlas. Over the
span of 15 years we were able to take advantage of new imaging
technologies such as OCT as well as faster computer systems
allowing implementation of sophisticated algorithms in real time. In
2015 we integrated OCT imaging into the database and implemented
better search algorithms (Taibl and Sayegh, ARVO 2015). In 2016
we have started an new phase in the development of RetinAsk by
implementing simplified forms of computer-assisted diagnosis (CAD)
rooted in algorithms that map image features to language tags and
integrate features from multiple imaging modalities. We also allowed
for the front end to allow “superusers” to submit labeled images and
for the program to group them with other, similarly tagged, images.
Torsional Indirect Traumatic Optic Neuropathy (TITON):
Identifying Biomarkers of Trauma using Matrix Assisted Laser
Desorption/Ionization (MALDI)
Kirstin Jones1, Randolph D. Glickman2, Brooke I. Asemota3,
Matthew A. Reilly1. 1Biomedical Engineering, UTSA, San Antonio,
TX; 2UTHSCSA, San Antonio, TX; 3St. Louis University School of
Medicine, St. Louis, MO.
Purpose: Current treatments for traumatic optic neuropathy (TON)
are largely ineffective and have adverse side effects. Development
of more effective treatments is hindered by a lack of understanding
of the molecular mechanisms underlying TON, as well as the lack
of a relevant animal model. To address these limitations, we have
developed a realistic rat model [Asemota et al., ARVO 2014], in
which we are investigating the pathophysiological and molecular
responses to torsional indirect TON (TITON). Earlier work from our
group, using matrix assisted laser desorption/ionization imaging mass
spectrometry (MALDI IMS), suggested that the pattern of protein
expression was modified in the optic nerve of TITON rats, compared
to controls [Glickman et al., ARVO 2014]. In the current work, we
tested the hypothesis that the induction of torsional indirect TON
(TITON) alters the expression of specific proteins in the eyes and
optic nerves of the rat model.
Methods: The eyes from the previous TITON study, which
induced torsional indirect TON using a robot, were used to test our
hypothesis. The subjects were adult female Sprague-Dawley rats
whose eyes were dissected seven days after injury. They were then
sectioned and prepared through a series of washes and sublimation
to be tested using MALDI IMS imaging. The mass spectra produced
by MALDI from control (n=2) and TITON optic nerves (n=2) were
analyzed for differences in protein expression and localization.
Results: Preliminary mass spectra show a decrease in the intensity of
protein peaks of the traumatized eyes compared to the control eyes.
This suggests that TON down-regulates protein expression for at
least seven days post injury. It is also suggested that changes are seen
mostly in the retina and choroid of the eyes. Further testing is likely
to confirm these findings and we are also optimistic that we will
begin to see changes in the optic nerve as well.
Conclusions: Our results are consistent with the hypothesis that TON
produces a qualitative change in the protein expression in the rat optic
nerve following TITON. Further analysis is underway to identify
specific proteins whose expression is related to trauma. Identification
will ultimately lead to improved understanding of the molecular
mechanisms governing TON, and possibly to the development of
more effective trauma therapies.
Commercial Relationships: Kirstin Jones;
Randolph D. Glickman, None; Brooke I. Asemota, None;
Matthew A. Reilly, None
Support: DOD Grant
Thickness of the macula, retinal nerve fiber layer, and ganglion
cell-inner plexiform layer in branch retinal vein occlusion :
the repeatability study of spectral domain optical coherence
tomography
Min Su Kim1, Hyung bin Lim1, Kyung sup Shin1, Jung Yeul Kim1, 2.
1
Ophthalmology, Chungnam National University College of
Medicine, Daejeon, Korea (the Republic of); 2R & D Division,
Chungnam Natl Univ Hosp, Daejeon, Korea (the Republic of).
Purpose: We performed a prospective, observational study to analyze
the repeatability of measurements of the thicknesses of the macula,
retinal nerve fiber layer (RNFL), and ganglion cell inner plexiform
layer (GCIPL) using spectral domain optical coherence tomography
(SD-OCT) in branch retinal vein occlusion (BRVO).
Methods: Patients diagnosed with BRVO were analyzed
prospectively. An experienced examiner obtained two consecutive
measurements from a macular cube 512 × 128 and optic disc cube
200 x 200 scans using SD-OCT. Thickness of central macula, RNFL,
and GCIPL of affected eyes with macular edema (before treatment),
without macular edema (after treatment), and opposite normal fellow
eyes were measured. Also, the repeatability of measurements was
evaluated with intraclass correlation coefficient (ICC).
Results: The average thickness of the central macula, RNFL, and
GCIPL was 411.35μm, 104.09μm, 51.03μm, respectively, in the
affected eyes with macular edema, 249.35μm, 94.56μm, 81.32μm
in the affected eyes without macular edema. The average GCIPL
was statistically significant thinner in the patients with macular
edema than without macular edema(p<0.05). The ICCs of the central
macula, RNFL, and GCIPL were 0.978, 0.919, 0.789, respectively,
in the affected eyes with macular edema, 0.999. 0.975, 0.928 in the
affected eyes without macular edema, showing high repeatability of
GCIPL after treatment. The average thickness of the central macula,
RNFL, and GCIPL was 253.12μm, 96.47μm, 85.09μm, respectively,
and the ICCs were 0.996, 0.995, 0.994 in the opposite normal fellow
eyes.
Conclusions: The macular contour change with the macular edema
in BRVO results in low repeatability and tendency to be measured
thinner in GCIPL thickness using SD-OCT. However, the average
thickness and repeatability of measurements of GCIPL increased
after resolution of macular edema. This can be explained by the
unstable gaze of the patient due to decreased visual acuity and autosegmentation error following changes in the macula.
Commercial Relationships: Min Su Kim, None; Hyung bin Lim;
Kyung sup Shin, None; Jung Yeul Kim, None
retinal vein occlusion (BRVO) and central retinal vein occlusion
(CRVO).
Methods: Baseline OCT scans and 3 monthly follow-up scans of
patients with macular edema secondary to CRVO (n=144) and BRVO
(n=96) were included. All patients received ranibizumab injections
for three months followed by a PRN regimen.
All OCT scans were automatically spatially aligned using the fovea
and optic disk center. Automatic segmentations of intraretinal cystoid
fluid and subretinal fluid (IRF/SRF) were created using convolutional
neural networks. From the segmentations, IRF and SRF volume was
computed for each section of the fovea centered ETDRS grid. Total
retinal thickness maps were automatically computed using the Iowa
layer segmentation reference algorithm.
To predict visual acuity (VA) and VA change to baseline for month 6,
elastic-net regularized linear regression models were trained on (1)
BCVA letter scores at baseline, (2) BCVA values of the loading phase
with their relative change from baseline, and (3) with additional
segmented imaging information from the initial 3 months.
To assess the predictive power of the models, a 5-fold crossvalidation setting was used. As error metrics, the mean absolute error
(MAE) in BCVA letters, and the predicted R2 were computed.
Results: Performance results of the prediction models are provided in
Table 1. Adding BCVA response during the loading phase increased
the predicted R2 of absolute/relative BCVA from 0.51/-0.01 to
0.78/0.54 for CRVO, and from 0.19/0.24 to 0.69/0.71 for BRVO.
Adding imaging features did neither reduce the MAE nor increase the
predicted R2.
Conclusions: The early response in BCVA score is a good predictor
of visual acuity outcomes in ranibizumab-treated CRVO and BRVO
eyes. Imaging information such as total retinal thickness and IRF/
SRF did not add an additional predictive value to the model.
Predicting future visual acuity outcomes from early morphologic
and functional response in anti-VEGF treated retinal vein
occlusion
Wolf-Dieter Vogl1, Sebastian M. Waldstein2, Bianca Gerendas2,
Thomas Schlegl1, Jing Wu2, Dominika Podkowinski2,
Ursula Schmidt -Erfurth2, Georg Langs1. 1Medical University
of Vienna, Christian Doppler Laboratory for Ophthalmic Image
Analysis, Department of Ophthalmology, Computational Imaging
Research Lab, Vienna, Austria; 2Medical University of Vienna,
Christian Doppler Laboratory for Ophthalmic Image Analysis,
Department of Ophthalmology, Vienna Reading Center, Vienna,
Austria.
Purpose: To determine predictive factors in longitudinal OCT data
for visual acuity outcomes after 6 months in patients with branch
Commercial Relationships: Wolf-Dieter Vogl, None;
Sebastian M. Waldstein, Bayer Healthcare AG (Berlin,
Germany) (C), Novartis Pharma AG, (Basel, Switzerland) (C);
Bianca Gerendas, None; Thomas Schlegl, None; Jing Wu, None;
Dominika Podkowinski, None; Ursula Schmidt-Erfurth, Bayer
Healthcare AG (Berlin, Germany) (C), Alcon Laboratories, Inc. (Fort
Worth, TX) (C), Boehringer Ingelheim GmbH (Ingelheim, Germany)
(C), Novartis Pharma AG, (Basel, Switzerland) (C); Georg Langs
Support: Austrian Federal Ministry of Science, Research and
Economy; National Foundation for Research, Technology and
Development.
Evolution of Short-wavelength Autofluorescence changes in
Central Serous Chorioretinopathy over 12 months
Marta Zola3, Priyanka Sanghi2, Namritha Patrao3, Deepthy Menon3,
Philip Hykin1, Sobha Sivaprasad1. 1NIHR Moorfields Biomedical
Research Center, London, United Kingdom; 2University College
London, Institute of Ophthalmology, London, United Kingdom;
3
Moorfields Eye Hospital, London, United Kingdom.
Purpose: Short-wavelength autofluorescence (SW-AF) is a useful
tool to assess the integrity of the retinal pigment epithelial cells
(RPE). The health of the RPE in central serous chorioretinopathy
(CSC) is often assessed by SW-AF. However, changes of SW-AF
over time may help determine prognosis. In this study, we evaluated
the changes in SW-AF over 12 months in different forms of CSC.
Methods: We retrospectively reviewed SW-AF images of 263
eyes with acute, recurrent and chronic CSC over 12 months. SWAF patterns were broadly classified into normal, reduced signal
in areas of subretinal fluid (SRF), stippled hyperautofluorescent
in areas of past or current SRF, plaque like confluent absent
autofluorescence, granular hypoautofluorescence, stippled hyper- and
hypoautofluorescence and tracts. Dome shaped macula and optic disc
pits were excluded. Changes in proportion of each SW-AF pattern
over 12 months are reported.
Results: In acute CSC (n=23), the most common pattern was reduced
signal in areas of SRF (n=21, 91%), other patterns were normal. 22%
of these patients (n=5) converted to stippled hyperautofluorescence in
the same area over time. No eyes converted to hypoautofluorescence.
In recurrent and chronic CSR (n=240), the most frequent pattern
at baseline was a stippled hyper and hypoautofluorescent pattern
(n=148, 62%) followed by confluent granular hypoautofluorescence
(n=53, 22%), confluent plaque like hypoautofluorescence (n=33,
14%) and normal (n=6, 2%). Tracts were present in 53 eyes (22%).
At 12 months, 4 out of the 6 eyes (67%) with normal baseline AF
worsened to granular hypoautofluorescence. None of the other
patterns of AF changed over 12 months but the areas of hyper- and
hypoautofluorescence in the stippled variety changed over time with
15% of existing patterns progressing to involve larger areas in eyes
with recurrent and chronic CSC.
Conclusions: Acute CSC does not convert to hypoautofluorescent
patterns within 12 months of presentation. In chronic and recurrent
CSC the patterns of autofluorescence at baseline do not change over
12 months but worsening of the same pattern may be seen by 12
months.
Commercial Relationships: Marta Zola, None; Priyanka Sanghi,
None; Namritha Patrao, None; Deepthy Menon, None;
Philip Hykin, Novartis (C), Bayer (C), Allergan (F), Bayer (F),
Novartis (F), Allergan (C); Sobha Sivaprasad, Novartis (C), Bayer
(C), Allergan (F), Bayer (F), Novartis (F), Allergan (C)
New Public Retinal Image Database for Tortuosity Evaluation
Jeffrey C. Wigdahl1, Roberto Annunziata2, Laura Hughes3,
Shyamanga Borooah4, Alfredo Ruggeri1, Emanuele Trucco2.
1
Information Engineering, University of Padova, Padova, Italy;
2
Vampire Project, School of Science and Engineering (Computing),
University of Dundee, Dundee, United Kingdom; 3College of
Medicine and Veterinary Medicine, University of Edinburgh,
Edinbrugh, United Kingdom; 4Center for Regenerative Medicine,
University of Edinburgh, Edinburgh, United Kingdom.
Purpose: Tortuosity in retinal images can be used as a biomarker
in the detection of several systemic diseases, including diabetes and
hypertension. This work provides a new retinal image database to test
and compare tortuosity metrics at both the vessel and image level, as
well as a comparison of several of the popular methods for tortuosity
estimation on the dataset.
Methods: One macula-centered image was acquired for each eye
in 37 patients (74 images) at the University of Edinburgh using a
Canon non-mydriatic camera at 45° field of view. A total of 100
arteries and 100 veins were chosen and graded from the images. The
tortuosity of these vessel segments was graded as either absent, low
or high by two clinical specialists. Image-level tortuosity was also
graded on the same scale by a total of 5 specialists. The database
consists of the retinal images, vessel centerline points from chosen
vessels (used to reproduce the vessel path), and the ground truth.
Six previously developed tortuosity metrics were tested against a
representative subset of the database (25 arteries and 25 veins). These
metrics are the arch /chord ratio (DM), tortuosity density (TD), slope
chain coding (SCC), and two integral curvature measures (Tau3, 5).
Descriptions of algorithms have been reported previously (Lisowska
et al., EMBC 2014). Full dataset and results will be made available at
http://bioimlab.dei.unipd.it/Data%20Sets.htm
Results: Intergrader variability was calculated for the subset of
vessels. Cohen’s kappa for agreement was .73 for veins and .61 for
arteries. Tortuosity metrics were calculated and agreement between
metrics and the two graders can be seen in Table 1. Results show that
no one metric had the highest agreement simultaneously across veins,
arteries, and graders. The tortuosity density metric had the highest
average agreement across all categories.
Conclusions: This work provides a new public database for
tortuosity estimation including images, vessel segments, and ground
truth. Results on a subset of vessels suggests that a single tortuosity
metric has difficulty capturing the qualitative grading of clinicians.
Table 1. Agreement between metrics and Graders. Values represent
Cohen’s kappa and linear weighted kappa statistic.
Fig. 1. Examples of different tortuosity levels (0-2) by row from
the database. Red and Blue dots mark the start and end points of the
vessels. For the images above, both graders were in agreement on the
tortuosity grade.
Commercial Relationships: Jeffrey C. Wigdahl, None;
Roberto Annunziata; Laura Hughes, None; Shyamanga Borooah,
None; Alfredo Ruggeri, None; Emanuele Trucco, None
A retrospective, observational analysis of the impact of a
telemedicine system to identify ocular disease in patients with
diabetes mellitus screened in a non-eye care medical setting
Ingrid E. Zimmer-Galler1, Yvonne Chu2, Christina Y. Weng2,
Jose A. Martinez3, Sunil Gupta4. 1Clinical-Retina, Johns Hopkins
Medical Institutions, Baltimore, MD; 2Baylor College of Medicine,
Houston, TX; 3Austin Retina, Austin, TX; 4Retinal Specialty Institute,
Pensacola, FL.
Purpose: Currently, less than 50% of patients diagnosed with
diabetes mellitus conform to the recommendation to undergo an
annual retinal evaluation to detect diabetes-related eye complications.
This retrospective, observational analysis was designed to
better understand the impact of a telemedicine screening system
implemented in non-eye care medical facilities to provide evaluations
for diabetic retinopathy.
Methods: A total of 59,347 patients with a prior diagnosis of diabetes
were screened in primary care health settings over a 36 month
period utilizing a commercially available telemedicine system.
The system includes an automated non-mydriatic fundus camera,
telemedicine platform, centralized reading center and grading
platform and a secure Internet-based data transfer portal. The images
were reviewed at the reading center and a report with the findings
and recommendations for referral were returned to the primary care
physician.
Results: Of the imaged patients, 34,709 (58.5%) were noted to
have ocular pathology of which 23,803 (40.1%) and 7,146 (12.1%)
respectively had mild nonproliferative diabetic retinopathy and
moderate nonproliferative diabetic retinopathy to proliferative
diabetic retinopathy that was previously undetected. Additionally,
other major ocular pathology that was detected included: suspected
glaucoma (4,550, 7.7%), suspected cataract (1,279, 2.2%), macular
edema (5,517, 9.3%), suspected hypertensive retinopathy (3,443,
5.8%), and suspected age-related macular degeneration (1,372,
2.3%).
Conclusions: Telemedicine platforms can provide first line screening
for diabetic retinopathy in a non-eye care medical setting. A large
number of patients can be screened in a cost-effective manner. Other
ocular pathology can also be detected in patients being evaluated
for diabetic retinopathy. This telemedicine system has demonstrated
an ability to increase patient compliance with well-established
recommendations for eye examinations in patients with diabetes.
Timely identification of advanced diabetic retinopathy will allow
appropriate referral and treatment with the ultimate goal of reducing
vision loss from diabetes.
Commercial Relationships: Ingrid E. Zimmer-Galler, Johns
Hopkins Medical Institutions; Yvonne Chu, Baylor College of
Medicine; Christina Y. Weng, Baylor College of Medicine;
Jose A. Martinez, Intelligent Retinal Imaging Systems (C);
sunil gupta, Genetech (C), Intelligent Retinal Imaging Systems (I),
USRetina (I), Intelligent Retinal Imaging Systems (P), Regeneron
(C), Alcon (C), Allergan (C)
Quantitative Fundus Autofluorescence (qAF) in Diabetic Patients
Andrew Hsu, Meleha Ahmad, Theodore Smith. Ophthalmology, New
York University School of Medicine, New York, NY.
Purpose: qAF is a recently developed technique that allows retinal
pigment epithelium lipofuscin fluorescence distribution to be
measured using confocal scanning laser ophthalmoscopy (SLO).
Little is known quantitatively regarding lipofuscin AF in the retinas
of diabetics. A single recent study has suggested lower lipofuscin
levels in diabetic macular edema (Yoshitake et al., 2015). In this pilot
study, we compare qAF in a cohort of diabetics to healthy controls.
Methods: 11 pseudophakic diabetic patients and 7 healthy
pseudophakes with clear posterior capsules were recruited
prospectively at a large city hospital. Background medical and ocular
history was collected prior to imaging. Following pupil dilation,
spectral-domain optical coherence tomography (SD-OCT) images
were taken to identify retinal pathology. Patients with any retinal
pathology were excluded from the control group; the presence of
diabetic retinopathy was noted in the study group. qAF images
were obtained by Spectralis with an internal fluorescence reference
(Delori et al., 2011). Images were uploaded to Igor Pro software
(WaveMetrics) to generate qAF intensity maps calibrated to the
internal reference. Mean qAF was calculated in an annular region
surrounding the fovea for each patient (Greenberg et al., 2013) and
corrected for intraocular lens type (blue-blocker or clear) and pupil
diameter 4.5 to 6 mm. One patient was excluded due to poorly
dilating pupil.
Results: Mean age was 74.1 ± 9.7 years for diabetic patients and 73.6
± 14.2 years for controls (P=0.93). Demographic information will be
presented. Diabetic patients had significantly lower qAF than controls
(134.1 ± 36.1 AU vs. 172.8 ± 28.7 AU P=0.023). Two patients were
identified with background diabetic retinopathy, with similar qAF as
in diabetics without evidence of retinopathy (P=0.634).
Conclusions: We observed lower qAF in diabetic maculas,
perhaps due to slowing of the visual cycle and reduced fluorophore
metabolism in diabetes. Alternately, increased optical media density
(ODm) of the posterior capsule and vitreous in diabetes could have
reduced the measured qAF. Additional studies with dark adaptometry
and independent ODm measurements are indicated to explain these
apparent reductions in lipofuscin in the diabetic macula.
Sample qAF intensity map in a diabetic patient. Average qAF
was calculated in area indicated by white circles (73 AU prior to
correction, 82 AU post-correction).
Commercial Relationships: Andrew Hsu, None; Meleha Ahmad,
None; Theodore Smith, None
Support: Foundation Fighting Blindness grant #BR-CL-0612-0575NYU and NIH Grant #R01 EY015520
The prognostic effect of peripheral non-perfusion on macular
thickness and visual acuity in diabetic and venous occlusive
retinopathies
Razek Georges Coussa, Cyril Archambault, Mikel Mikhail,
Fares Antaki, David E. Lederer. Ophthalmology, McGill University,
Montreal, QC, Canada.
Purpose: To study the prognostic effect of peripheral retinal nonperfusion on macular thickness and visual acuity in diabetic and
venous occlusive retinopathies.
Methods: 156 and 53 treatment naive eyes with diabetic (88% NPDR
and 12% PDR) and venous occlusion (40% BRVO, 13% HRVO and
47% CRVO) retinopathies, respectively, were randomly selected
from our practice between August 2008 and August 2014. Wide-field
retinal funds photos and fluorescein angiography (FA) were obtained
on all patients using the Optos 200 Tx system (Optos plc, Scotland,
UK). The peripheral perfusion index (PI), which was defined as the
area beyond 45 degrees, was calculated using ImageJ 1.43 software.
Student t-tests and regression analyses were used to analyze the data.
Results: 47% of DR and 55% of RVO subjects showed significant
linear correlation between % peripheral and central non-perfusion.
Ignoring peripheral non-perfusion misses ≈50% of total nonperfusion. In NPDR, macular thickness and central non-perfusion
statistically affected VA. The average extent of peripheral nonperfusion was 2.2x fold larger in RVO compared to DR (8.8% RVO
vs. 3.9% DR). In RVO, the effect of peripheral non-perfusion on VA
was 2x folds significantly larger than that of macular thickness.
Conclusions: Peripheral non-perfusion sheds potential prognostic
insights in vaso-occlusive retinopathies. In the future, peripheral
non-perfusion could potentially be included in vaso-occlusive
retinopathies’ disease staging in order to enhance guided treatment
options.
Commercial Relationships: Razek Georges Coussa, None;
Cyril Archambault, None; Mikel Mikhail, None; Fares Antaki,
None; David E. Lederer, Alcon Inc (C), Novartis Pharmaceuticals
Inc. (C), Bayer Inc (C)
Automated Detection and Classification of Longitudinal Retinal
Changes Due to Microaneurysms for Diabetic Retinopathy
Screening
Kedir Adal1, 3, Peter van Etten2, Jose P. Martinez2, Kenneth Rouwen2,
Lucas J. van Vliet3, Koenraad A. Vermeer1. 1Rotterdam Ophthalmic
Inst, Rotterdam, Netherlands; 2Rotterdam Eye Hospital, Rotterdam,
Netherlands; 3Delft University of Technology, Delft, Netherlands.
Purpose: To present and evaluate a multi-stage automated framework
for the detection and classification of longitudinal retinal changes due
to microaenurysms (MAs) for diabetic retinopathy (DR) screening.
Methods: Fundus image sets from one eye of each of 82 diabetic
patients who were screened for DR in 2012 and 2013 were used for
training (30 eyes) and testing (52 eyes) the framework. First, the
fundus image sets acquired during successive retinal examinations
were normalized for illumination variation and registered into
a common coordinate system (Adal et.al.,IOVS,2015). Second,
candidate spatio-temporal retinal change locations were extracted by
a novel multiscale Laplacian of Gaussian (LoG) algorithm. Third,
several intensity and shape descriptors were extracted from each
candidate region and subsequently used by a support vector machine
(SVM) to classify the region as an MA or a non-MA related retinal
change.
The fundus mosaics of each eye were independently annotated
by two graders for MA related retinal changes between the two
screening time-points. Different ways of combining the two graders’
annotations were used to define a ground truth.
Results: The performance of the proposed framework was evaluated
on the image sets of 52 eyes. The system achieves a sensitivity of
90% in finding MA related changes marked by both graders at an
average of 5 false change detections per image set (fig 1). Some of
these false detections relate to other dark-red lesions that resemble
MAs (fig 2).
Conclusions: The system is able to detect retinal changes, including
those that are visually difficult to detect on the color fundus images.
The detected MA related changes can be used as a biomarker for
objective assessment of DR progression, such as the MA turnover
rate, as well as for more efficient human grading by highlighting DRrelated changes since the previous exam.
Free-response ROC curves of the proposed framework for the
detection and classification of MA related retinal changes.
Examples of color and normalized fundus image patches showing
the automatically detected MA related change locations (blue circles)
between the baseline (left) and follow-up (right) DR checkups.
The yellow arrows indicate a change due to a hemorrhage and thus
considered as false alarm. The green arrows indicate locations that
were annotated by one grader, all other locations were annotated by
both graders.
Commercial Relationships: Kedir Adal, None; Peter van Etten,
None; Jose P. Martinez, None; Kenneth Rouwen; Lucas J.
van Vliet, None; Koenraad A. Vermeer, None
Support: Stichting Achmea Gezondheidszorg, Stichting Coolsingel
Combining Medical Data and Fundus Images to Detect Eye
Diseases in Patients with Diabetes
Carla Agurto Rios1, Sheila C. Nemeth1, Gilberto Zamora1,
Wendall Bauman2, Peter Soliz1, E Simon Barriga1. 1VisionQuest
Biomedical LLC, Albuquerque, NM; 2Retina Institute of South Texas,
San Antonio, TX.
Purpose: To integrate retinal images with other personal medical
information, to develop a comprehensive eye evaluation algorithm
that detects diabetic retinopathy (DR), age related macular
degeneration (AMD), and risk for glaucoma in patients with diabetes,
and also to show that this automatic detection can be improved when
considering other medical information.
Methods: Studies have shown that diabetic patients are twice
likely to have glaucoma and are at risk for AMD. Thus, screening
programs that integrate glaucoma and AMD screening in their DR
examinations are highly cost effective. Based in our previous work
in automatic detection of DR, we expanded our system to detect
these other two eye diseases. Our algorithm uses statistical features
from retinal images and other medical information such as age,
gender information, duration of diabetes, presence of hypertension
and cataracts. Then, we combined these features using a partial least
squares (PLS) classifier.
To test our method, we used data from N=85subjects with diabetes.
Of these, N=24 eyes had AMD, N=29 had DR, N=25 optic disc
indicators of glaucoma (glaucoma suspect), and N=26 were controls,
i.e., diabetics that did not present with any eye disease. In addition,
we obtained medical history information such as age, diabetes
duration, among others. The images were acquired using Canon
retinal cameras. For each eye, optic disc-centered (field 1) and foveacentered (field 2) images were used.
Results: Table 1 shows the classification results in terms of AUC and
sensitivity/specificity. Marked improvement was achieved for AMD
classification, going from an AUC of 0.77 to 0.81 by adding the
medical information. For glaucoma, we achieved a 3% improvement
when adding medical features. For DR, image features were sufficient
to achieve excellent classification accuracy (0.93). In general, the
classification of combined eye disease also improved when the
medical features are added. In this case, a 2% increase in AUC
corresponds to 5% sensitivity from 84% to 89%, while the specificity
remains in 64 %.
Conclusions: We present a new automatic system for the detection
of the three main causes of visual impairment in the US: Glaucoma,
AMD and DR. Our results show that the combination of retinal
features with other patient health information has the potential to
increase the performance of algorithms for automatic detection of eye
diseases.
Table 1 Preliminary results
Commercial Relationships: Carla Agurto Rios, VisionQuest
Biomedical LLC; Sheila C. Nemeth, VisionQuest Biomedical LLC;
Gilberto Zamora, VisionQuest Biomedical LLC; Wendall Bauman,
Retina Institute of South Texas (I); Peter Soliz, VisionQuest
Biomedical LLC (I); E Simon Barriga, VisionQuest Biomedical
LLC
Support: NEI Grant R44EY018280-05S1
Fluorescence Lifetime Imaging Ophthalmoscopy in Diabetic
Retinopathy
Martin Hammer1, 2, Johanna Schmidt1, Sven Peters1, Lydia Sauer1,
Nicole Müller3, Matthias Klemm4, Regine Augsten1, Daniel Meller1.
1
Ophthalmology, University Hospital Jena, Jena, Germany; 2Center
for Medical Optics and Photonics, Univ. of Jena, Jena, Germany;
3
internal Medicine, University Hospital Jena, Jena, Germany;
4
Technical University, Ilmenau, Germany.
Purpose: To investigate metabolic byproducts from diabetic
retinopathy by fundus autofluorescence (FAF) lifetime imaging
(FLIO) and to discriminate patients from healthy controls.
Methods: 33 patients suffering from non-proliferative diabetic
retinopathy (NPDR) and 28 age-matched controls were subjected to
retinal fluorescence lifetime measurement using FLIO (Heidelberg
Engineering, Heidelberg, Germany). Autofluorescence of a 30 degree
retinal field was excited by picosecond laser pulses (75 ps FWHM)
at 468 nm and fluorescence decay over time was recorded in two
spectral channels (498-560nm and 560-720nm). FAF decays were
approximated by a series of three exponential functions (least square
fit) in each pixel of the FAF image resulting in three lifetimes (τ1-τ3),
and three respective amplitudes (a1-a3). Main outcome measure
was the amplitude-weighted mean lifetime τm. All parameters were
averaged over the subfields of the standard ETDRS-grid centered at
the macula.
Results: FAF lifetimes τ1-τ3 as well as tm were extended in the
patients compared to controls (Man-Whitney-U-test, p<0.05, fig.
1). This was found predominantly in the short-wavelength channel
for the inner ring of the ETDRS-grid: τm=333±141ps vs. 220±79ps
(p=0.001). Statistically independent FLIO-parameters were combined
in a logistic regression model. A sensitivity of 90.1% and a specificity
of 71.4% were found for the discrimination of NPDR-patients (area
under the ROC-curve: 0.865).
Conclusions: Independent from other clinical signs of NPDR,
FLIO can hint on diabetic alterations at the posterior pole of the
eye. The extension of fluorescence lifetimes, predominantly at short
wavelengths, may be explained by the accumulation of advanced
glycation end products which showed long lifetimes in previous invitro investigations peaking at 500 nm. FLIO should be considered
as a new imaging technique capable to detect specific fluorophores at
the fundus with possible pathognomonic implication.
Color-coded mean FAF lifetime (τm) images of a control subject
(left) and a NPDR patient (right).
Commercial Relationships: Martin Hammer, None;
Johanna Schmidt, None; Sven Peters, None; Lydia Sauer, None;
Nicole Müller, None; Matthias Klemm, None; Regine Augsten,
None; Daniel Meller, None
Can Retinal Vascular Geometry predict future progression of
Diabetic Retinopathy?
Maged Habib2, Bashir Al-Diri1, Roxanne R. Crosby-Nwaobi3,
Sobha Sivaprasad3, David Steel2. 1Computer Science, Lincoln
University, Lincoln, United Kingdom; 2Sunderland Eye Infirmary,
Sunderland, United Kingdom; 3Moorfields Eye Hospital, London,
United Kingdom.
Purpose: Geometrical changes in the retinal vascular network have
been reported to be associated with different stages of diabetic
retinopathy (DR). Previous studies utilising semi-automated vascular
analysis programmes demonstrated potential role of these vascular
changes in diagnosis and monitoring disease progression. This study
was conducted to evaluate the predictive role of retinal vascular
analysis in future progression to proliferative diabetic retinopathy
(PDR) using of a novel fully automated vascular analysis tool.
Methods: Experimental study using baseline images of 15 diabetic
subjects that showed future progression to PDR (progressors)
as compared to 28 subjects that showed no signs of retinopathy
progression (non-progressors) over the same period of time. Retinal
coloured images that preceded the onset of any DR in both groups
were used for this study. Geometrical parameters such as area ratios,
bifurcation angles, and optimality ratios and deviation factors were
estimated and compared for both groups.
Results: The branching angle of larger daughter-vessel branch at
a vascular bifurcation at baseline was significantly wider in the
progressors group as compared to non-progressors (p=0.023).
Significant changes in Optimality parameter as well as optimality
deviation were also noted between both groups (p= 0.035 and 0.037
respectively). There were no differences noted in other parameters
Conclusions: Geometrical vascular changes at baseline before the
development of any DR can be detected utilising a fully automated
system that demonstrate abnormal retinal vascular networks’ patterns
that can be at risk for future development of PDR. Such findings
may be used in clinical practice for individualising patient’s care and
planning monitoring intervals based on their estimated progression
risks.
Commercial Relationships: Maged Habib, None; Bashir Al-Diri,
None; Roxanne R. Crosby-Nwaobi; Sobha Sivaprasad, None;
David Steel, None
Atypical vascularization of the foveal avascular zone in the
human macula
Delia DeBuc1, Jing Tian1, Thalmon R. Campagnoli1, WenHsiang Lee1, Hong Jiang1, Jianhua Wang1, Sagi -. Reuven2,
Amiram Grinvald5, William E. Smiddy1, Gabor M. Somfai3, 4.
1
Ophthalmology, University of Miami, Miami, FL; 2Optical Imaging,
Ltd, Rehovot, Israel; 3Retinology Unit, Pallas Kliniken, Olten,
Switzerland; 4Ophthalmology, Semmelweis University, Budapest,
Hungary; 5Weizmann Institute of Science, Rehovot, Israel.
Purpose: To present different patterns of atypical foveal
vascularization observed in healthy individuals and in patients with a
variety of ocular and systemic diseases during studies with the retinal
functional imager (RFI, Optical Imaging Ltd., Rehovot, Israel).
Methods: We analyzed the study data of subjects enrolled in
studies involving healthy controls and patients with diabetes and no
retinopathy, diabetes and mild non-proliferative diabetic retinopathy,
central and branch retinal vein occlusion and multiple sclerosis (MS).
All subjects underwent retinal blood flow imaging and capillary
perfusion mapping using the RFI. We were aiming to identify
abnormally crossing capillaries in the region of the foveal avascular
zone (FAZ).
Results: Abnormal FAZ pattern was present in 8 (4%) of 224 eyes
of 6 (6%) of 107 patients (36 ±8 years old). Those with atypical FAZ
patterns were a type 2 diabetic female without diabetic retinopathy,
three seemingly healthy males, one female with MS and one male
with cystoid macular edema due to central retinal vein occlusion.
Retinal circulation and capillary patterns showed abnormal FAZ in
both healthy and pathological patients (see Fig. 1). Specifically, all
retinas contained capillaries crossing the fovea in various patterns.
The blood flow pattern of the atypical perifoveal capillaries was
indistinguishable from that of the more peripheral capillaries. In
adition, the size of the FAZ was variable and smaller in area than in
the normal healthy population as reported in the literature.
Conclusions: Recent advances in in vivo optical imaging
technologies have enabled better visualization of the perifoveal
capillary bed, showing vasculature that is not apparent by other
means (e.g. fundus photo). The findings of this study have
demonstrated patterns of retinal circulation in the FAZ where
capillaries are usually absent. These patterns are present in apparently
healthy patients as well as in patients with various ocular and
systemic diseases. Further studies should investigate whether atypical
perifoveal capillary patterns in the FAZ are induced by disease,
degeneration or have a developmental explanation. Such studies will
better inform our understanding of the factors that may determine
atypical capillary occurrence, significance and their relationship to
macular function.
Commercial Relationships: Delia DeBuc, Optical Imaging,
Ltd (C); Jing Tian, None; Thalmon R. Campagnoli, None;
Wen -Hsiang Lee, None; Hong Jiang, None; Jianhua Wang,
Optical Imaging, Ltd (C); Sagi -. Reuven, Optical Imaging, Ltd,
Optical Imaging, Ltd (I); Amiram Grinvald, Optical Imaging,
Ltd (C), US Patent 6,588,901 (P), Optical Imaging, Ltd (I);
William E. Smiddy, None; Gabor M. Somfai
Support: This study was supported in part by a NIH Grant No.
NIH R01EY020607, a NIH R01EY020607S, a NIH Center Grant
No. P30-EY014801, by an unrestricted grant to the University of
Miami from Research to Pre- vent Blindness, Inc., and by an Eotvos
Scholarship of the Hungarian Scholarship Fund.
Optical Coherence Tomography Minimum Intensity as an
objective measure for the detection of hydroxychloroquine
toxicity
Ali M. Allahdina1, Paul F. Stetson2, Wai T. Wong1, Emily Y. Chew1,
Catherine A. Cukras1. 1National Eye Institute, NIH, Bethesda, MD;
2
Research and Development, Carl Zeiss Meditec, Dublin, CA.
Purpose: Spectral domain Optical Coherence Tomography (SDOCT) has been previously shown to be a useful modality in the
detection of hydroxychloroquine toxicity both for qualitative
inspection and quantitative ananlysis. OCT Minimum Intensity
(MI) provides a different analysis of the OCT data to visualize
photoreceptor disruption. We compare OCT MI analysis to multifocal
electroretinography (mfERG) reference testing in participants with
and without toxicity.
Methods: Fifty-seven study participants (91.2% female; mean age,
55.7±10.4 years; mean duration of hydroxychloroquine treatment,
15.0±7.5 years) were divided into toxicity affected (n=19) and
unaffected (n=38) groups using objective mfERG criteria. A CirrusHD system was used to obtain macular 512x128 cube scans and were
analyzed using the OCT MI algorithm, a novel quantitative technique
developed by Carl Zeiss Meditec, Inc. (Dublin, CA). The OCT MI
results were analyzed in each ETDRS subfield and data for one eye of
each participant was used in a t-test analysis comparing the affected
and unaffected groups. A receiver operating characteristic (ROC)
curve was plotted and area under the curve (AUC) was calculated
for each subfield to assess the sensitivity and specificity of the OCT
MI to discriminate between presence and absence of HCQ retinal
toxicity.
Results: The mean of the median OCT MI values measured in all
9 ETDRS subfields were significantly higher in the affected group
compared with those of the unaffected group (median difference in
MI > 3.50, P < 0.005 for all comparisons). The subfields with the
greatest difference in the median OCT MI values between affected
and unaffected patients were the Inner Inferior subfields (median
difference MI >10.00; P<0.0001). ROC analysis of Inner Inferior
subfields showed high sensitivity and high specificity (AUC= 0.997;
Sensitivity=100%, Specificity= 89.5%, with MI cutoff= 54.3).
Conclusions: Analysis of the OCT MI, especially in the Inner
Inferior subfield, is an objective anatomical measure that appears
to demonstrate clinically useful sensitivity and specificity for the
detection of hydroxychloroquine toxicity as identified by mfERG.
OCT is a widely accessible imaging modality and incorporating
OCT MI analysis to patient evaluation may aid in the detection of
hydroxychloroquine-induced retinal toxicity.
Commercial Relationships: Ali M. Allahdina, None;
Paul F. Stetson, Carl Zeiss Meditec, Carl Zeiss Meditec (P);
Wai T. Wong, None; Emily Y. Chew; Catherine A. Cukras, None
Clinical Trial: NCT01145196
Quantitative fundus autofluorescence in patients treated with
hydroxychloroquine
Francesco Viola1, 2, Maura Di Nicola3, 2, Eleonora Benatti3, 2,
Elena Tabacchi2, Riccardo Clerici3, Alessandro Santaniello2,
Alessandro Invernizzi2. 1Department of Clinical Sciences and
Community Health, University of Milan, Milan, Italy; 2Fondazione
IRCCS Cà Granda Ospedale Maggiore Policlinico, Milan, Italy;
3
University of Milan, Milan, Italy.
Purpose: To quantify fundus autofluorescence (qAF) in patients
treated with hydroxyichloroquine (HCQ) and to find a possible
correlation between qAF variation and cumulative drug intake.
Methods: 50 patients treated with HCQ were enrolled in the study
with a consecutive stratified strategy according to the cumulative
drug intake and divided into 5 risk classes for HCQ retinopathy. 10
healthy age and sex-matched volunteers were recruited as control
group.
All the subjects underwent a complete ophthalmologic examination
including best corrected visual acuity, slit lamp, fundus photography
(FP), Spectral Domain-Optical Coherence Tomography (SDOCT), fundus blue autofluorescence (BAF) and qAF (Hardware
and Software by Heidelberg Engineering - not for clinical use). FP,
SD-OCT and BAF images were analyzed by trained operators for
the presence of visible alterations referable to HCQ toxicity. qAF
values were automatically calculated for single subfields of the Delori
pattern overlay centered onto the fovea. Mean qAF values for each
concentric ring of the pattern were then obtained. qAF variations
among the classes were analyzed by ANOVA.
Results: FP, SD-OCT and BAF didn’t show any detectable alteration
in the subjects treated with HCQ neither in low risk classes (I to
III = cumulative dose < 1000 g) nor in high risk classes (IV-V
= cumulative dose >1000). On the contrary qAF values showed
an increasing trend from class I to V. In particular, significantly
higher values were noted in high risk classes (IV-V) as compared to
healthy controls (both p<0,01) regardless the considered ring of the
Delori pattern. The innermost ring showed the highest significance
(p<0,001). Foveal subfield didn’t show any significant difference
among the classes in qAF values.
Conclusions: qAF values seem to increase in patients treated
with HCQ before qualitative changes on FP, BAF or SD-OCT
can be detected. qAF variations show a direct correlation with the
cumulative drug intake, with the patients at high risk for developing
HCQ retinopathy having significantly higher values as compared to
healthy controls. qAF could represent a possible screening tool for
the early diagnosis of HCQ retinopathy.
between vegetarians and non-vegetarians in the older population. The
average levels of autofluorescence are included in Table 1.
Conclusions: In this cohort group ages 40-55 yrs, vegetarians
had statistically significant lower levels of FAF than nonvegetarians (p=6.59x10-10*) with a 34% increase in levels of mean
autofluorescence luminance in P seen in non-vegetarians. Results
from our pilot study showed a 23% increase in levels of mean
autofluoresnce luminance in P from vegetarians ages 20-15 yrs to
non-vegetarians in the same age range. When compared to the pilot
study, this follow-up study supports our initial findings and also
identifies the increase in autofluorescence seen with increased age.
Further studies will include analysis of FAF quantitative scores in
healthy and non-healthy RPE as well as gender specific analysis.
Commercial Relationships: Francesco Viola, None; Maura Di
Nicola, None; Eleonora Benatti, None; Elena Tabacchi,
None; Riccardo Clerici, None; Alessandro Santaniello, None;
Alessandro Invernizzi, None
Lipofuscin RPE Imaging in Vegetarians and Non-vegetarians:
Dietary and Age Effects
Pooja A. Padgaonkar, Sumana S. Kommana, Lesley Wu,
Nicole Mendez, Bernard C. Szirth, Albert S. Khouri. Rutgers-New
Jersey Medical School, Newark, NJ.
Purpose: Identify the relationship between vegetarian vs. nonvegetarian diet and lipofuscin build up in the retinal-pigmented
epithelium (RPE) of an adult South Asian population (age 40-55
yrs) using fundus autofluorescence (FAF) imaging and compare
results to those from a similar previous study comparing lipofuscin
concentration in ages 20-25 yrs.
Methods: Sixty two subjects underwent FAF imaging (mean 47 yrs,
SD 4) using a Canon CR-2 Plus AF retinal camera with an 18 Mp
CMOS sensor fitted with an exciter (535-585 nm wide band) and
barrier filter (605-715nm). All monochromatic images were captured
without mydriatic agents at a flash setting of 300 w/sec with an
angle of 45° and corrected for auto-exposure (Adobe Photoshop V
7.02). Quantitative autofluorescence measurements of a 35.5 mm2
rectangle in the paracentral retina (P) were used as a measure of
lipofuscin accumulation. Mean, SD and T-tests allowed comparison
of autofluorescence in vegetarians vs. non-vegetarians ages 40-55 yrs
as well as for comparison of vegetarians and non-vegetarians in this
study (age 40-55 yrs) to their corresponding group in our pilot study
(age 20-25 yrs).
Results: Sixty-two eyes were analyzed (47% females, all South
Asian decent, 1.6% smokers); of those, 48% percent were vegetarian,
52% non-vegetarian (defined as eating meat more than ½ their life).
Autofluorescence scores in P were significantly different (p < 0.05)
Table 1. T-test results from comparison of age (2 tail, 3) and FAF
score in P (1 tail, 3) seen in vegetarians vs. non-vegetarians ages 4055 yrs (*p<0.05).
Commercial Relationships: Pooja A. Padgaonkar, None;
Sumana S. Kommana, None; Lesley Wu, None; Nicole Mendez,
None; Bernard C. Szirth, None; Albert S. Khouri, None
Retinal and choroidal features in active posterior uveitis
and panuveitis assessed by swept-source optical coherence
tomography
Alfredo Adan Civera, Jessica Matas, Angels De Pouplana,
Victor Llorenç, Marina Mesquida, Anna Sala, Maria Teresa Sainz de
la Maza, Javier Zarranz-Ventura. Ophthalmology, Hospital Clinic,
Barcelona, Spain.
Purpose: To evaluate the retinal and choroidal changes observed
in patients with uveitis assessed by swept source optical coherence
tomography (SS-OCT, Atlantis DRI OCT-1, Topcon, Japan).
Methods: Single centre consecutive case series. SS-OCT images
were qualitatively and quantitatively assessed after being manually
corrected for segmentation errors. Qualitative analysis included:
1) presence of vitreomacular traction (VMT), 2) disruption of the
ellipsoid line (EL), 3) presence and location of hyperreflective dots
(HRD), 4) presence of subretinal and intraretinal fluid, 5) integrity
of external limiting membrane (ELM), and 6) retinal pigment
epithelium detachment with or without fluid, and 7) identification
of the suprachoroidal space. Quantative analysis included a) retinal
and b) choroidal thickness in the macula (early treatment of diabetic
retinopathy study ETDRS grid total macular circle) and fovea (central
subfield of the ETDSR grid).
Results: Twenty-five uveitis eyes (16 patients) were included in
the study. 92% were non-infectious uveitis and 8% infectious, and
according to the site of inflammation, uveitis was classified as
posterior in 60.8% and panuveitis in 39.1% as per the Standardization
of Uveitis Nomenclature criteria. VMT was observed in 16%, with
disruption of the EL in 52% and presence of HRD in 24% of study
eyes (83,3% in the outer retina and 16,6% in the inner retina).
Subretinal and intraretinal fluid was detected in 28% and 40% of the
study cohort, whereas ELM disruption was seen in 32% and RPE
elevation in 36% (without subretinal fluid in 77,7% and with fluid
in 22,2%) of study eyes. The suprachoroidal space was identified in
only 12% of the cases. Mean retinal and choroidal thickness in the
total macular circle was 301.6±44.8mm and 295.8±98.4mm, and in
the fovea 273.1±89.6 and 296.9±118.8mm respectively.
Conclusions: SS-OCT allows adequate identification of retinal and
choroidal features in posterior and panuveitis eyes. The longer wavelength of the laser source permits accurate delineation of deep retinal
layers. However, the suprachoroidal space was observed only in a
quarter of the study eyes. Future studies are required to elucidate the
role of SS-OCT in the assessment of choroidal structures in uveitis.
Commercial Relationships: Alfredo Adan Civera, None;
Jessica Matas, None; Angels De Pouplana; Victor Llorenç, None;
Marina Mesquida, None; Anna Sala, None; Maria Teresa Sainz de
la Maza, None; Javier Zarranz-Ventura, None
CORRELATION IN RETINAL NERVE FIBER LAYER
THICKNESS IN UVEITIS AND HEALTHY EYES USING
SCANNING LASER POLARIMETRY AND OPTICAL
COHERENCE TOMOGRAPHY
David Bellocq, Laurent Kodjikian, Philippe Denis. Ophthalmology,
Croix Rousse University Hospital, LYON, France.
Purpose: To evaluate the correlation of retinal nerve fiber layer
(RNFL) thickness measured by spectral domain optical coherence
tomography (SD-OCT) and scanning laser polarimetry (SLP) in
uveitis eyes compared to healthy eyes
Methods: Descriptive, observational, prospective, consecutive,
controlled, monocenter case series have been conducted from
May to October 2015. Clinical characterizes, best-corrected visual
acuity (BCVA), intra-ocular pressure (IOP), retinal nerve fiber
layer (RNFL) thickness measurement with spectral domain optical
coherence tomography (SD OCT) and scanning laser polarimetry
(SLP) using GDx variable corneal compensation (GDx VCC)
have been performed for each patient. An evaluation of anterior
chamber inflammation with laser flare cell meter has also been made.
Correlation between SD-OCT and GDx VCC RNFL measurement
was evaluated with linear regression analysis.
Results: Fifty-four patients have been included and divided in two
groups: 50 healthy eyes of 29 patients and 42 uveitic eyes of 25
patients. The mean RNFL thickness was 98.08 (standard deviation
8.42) and 113.21 (20.53) microns in the healthy group and the uveitic
group by SD-OCT (p<0,001) and 56.43 (5.24) and 58.77 (6.67)
microns by GDx VCC (p=0.078) respectively. The total average
RNFL thickness correlated highly when measured using SD-OCT and
GDX (r=0.48, p<0.001) in healthy eyes but not in uveitic eyes (r=0.2,
p=0.19).
Conclusions: RNFL thickness was significantly increased using
SD-OCT in active uveitis compared to GDx VCC. RNFL thickness
measurement seems not to be correlated between these two
techniques in active uveitic eyes. Discrepancy between the results
obtained with SD-OCT and GDx VCC may suggest that accurate
measurement of RNFL may associate the use of the two techniques
for these patients.
Commercial Relationships: David Bellocq, None;
Laurent Kodjikian, None; Philippe Denis, None
Ultra-Widefield Fluorescein Angiography in Intermediate Uveitis
Wipada Laovirojjanakul, Nisha Acharya, John A. Gonzales. Uveitis,
Francis I. Proctor Foundation, UCSF, San Francisco, CA.
Purpose: Ultra-widefield fluorescein angiography (UWFFA)
in intermediate uveitis often reveals vascular leakage that is not
clinically detectable. We describe patterns of vascular leakage using
UWFFA and correlate these findings with clinical manifestations of
inflammation in intermediate uveitis.
Methods: We performed a retrospective review of 24 patients with
intermediate uveitis who had UWFFA performed at a single tertiary
referral center, The Francis I. Proctor Foundation, between July 2014
and July 2015. The main outcome was the pattern of retinal vascular
leakage on UWFFA, which was classified as being anterior or
posterior to the equator, or both. Visual acuity (VA), inflammation on
exam and CME was also assessed. The association between activity
level and location of leakage on UWFFA was assessed by using the
chi-squared test.
Results: A total of 41 eyes from 24 patients were included, of which
31 eyes (76%) displayed active inflammation and 10 eyes displayed
inactive inflammation (24%). We identified 2 patterns of vascular
leakage: 1) anterior to the equator (12 eyes, 29%) and 2) posterior
and anterior to the equator (26 eyes, 63%). No eyes exhibited
leakage posterior to the equator alone. Of the 31 eyes with active
inflammation 21 (77%) had posterior and anterior leakage while
7 eyes (23%) had leakage anterior to the equator alone. Of the 10
clinically inactive eyes 2 (20%) had posterior and anterior leakage,
5 eyes (50%) had anterior leakage alone, and 3 eyes (30%) had no
leakage. Of the 10 eyes that were clinically inactive, 2 eyes (20%)
exhibited posterior and anterior leakage There 5 eyes (50%) with
anterior leakage alone. All of these eyes had VA better than 20/40.
Three eyes (30%) did not have any leakage on UWFFA. We found a
statistically significant difference in the number of the patients with
active inflammation demonstrating leakage on UWFFA compared to
patients with inactive leakage (p = 0.012).
Conclusions: We identify leakage on UWFFA not appreciated on
clinical examination. Given that the current classification criteria
considers intermediate uveitis to have inflammation in the vitreous
which may be accompanied by leakage of the peripheral retinal
vasculature, consideration for revision to the nomenclature for
intermediate uveitis exhibiting vascular leakage posterior to the
equator should be considered. The clinical significance of leakage
anterior to the equator remains to be determined.
Commercial Relationships: Wipada Laovirojjanakul, None;
Nisha Acharya, None; John A. Gonzales, None
Comparison of Common Radiographic Techniques in Identifying
Intraoperatively “Lost” Surgical Needles and Vitrectomy Trocars
Yigit Akduman2, Jason M. Newman3, Maria Zulfiqar1,
Ece I. Akduman1. 1Saint Louis University, St. Louis, MO; 2Clayton
High School, St. Louis, MO; 3SureVision, St. Louis, MO.
Purpose: Obtaining x-rays is a common part of operating room
protocol for intraoperatively lost needles, trocars and sclerotomy
plugs. However, small needles and trocars may escape the detection
threshold of routine x-rays. This practice pattern exposes patients
and personnel to unnecessary radiation, prolongs operating room
turn-over time and may provide false assurance to the surgeon due
to the high rate of false negatives. Extended surgical and anesthesia
time may also increase patient morbidity and mortality. We compared
various needle, trocar and scleral plug sizes to determine the
minimum threshold that could be identified by x-ray, fluoroscopy, and
CT scan in an effort to guide best practice for intraoperatively lost
needles, trocars and scleral plugs.
Methods: We placed a variety of potentially lost surgical items
around the orbit of a phantom human skull. These included suture
needles ranging in length from 3.8mm to 40mm with correlating
suture sizes of 10-0 to 2-0, 23 gauge vitrectomy trocars, as well
as 19 and 20 gauge sclerotomy plugs. We then performed x-ray,
fluoroscopy and CT scans using standard radiography doses and
techniques.
Results: X-ray and fluoroscopy were each able to identify needles
8 mm and larger (i.e. those from sutures that were ~6-0 in size or
larger) and 19g and 20g sclerotomy plugs. CT was able to identify all
needles, trocars and plugs tested. The 23g trocar was not visible on
x-ray. Needles overlying complex bony anatomy such as the inferior
orbital rim were less likely to be seen due to multiple overlapping
shadows in contrast to needles placed over simpler osseous
structures, such as the frontal bone.
Conclusions: X-ray and fluoroscopy are unable to reliably locate
small lost surgical items including needles smaller than 8 mm in
length (~6-0) or 23 gauge vitrectomy trocars. In these cases CT can
be done, which is able to identify lost surgical items of all tested
sizes. In cases where a lost surgical needle of 8mm in length or less
is strongly suspected to be in the tissue, a CT scan may be a more
appropriate technique. Our study was done for a sampling of suture
needle sizes. There is a wide variety of suture needle sizes and the
ability to identify them on imaging depends on their contrast to the
tissue in which they are superimposed. Therefore, a larger study is
recommended to validate our findings and help guide policy.
Commercial Relationships: Yigit Akduman, None;
Jason M. Newman, None; Maria Zulfiqar, None; Ece I. Akduman,
None

Session 325 Clinical Imaging

Transcription

Similar documents

Practice Brochure - Arizona Retina Institute

Madagascan tomato frog

Bionic Eyes - Hansma Lab

Retinal Arterial Macroaneurysm

TRANSPORTATION

Retinal Toxicity - Healthcare Professionals

Color Variability Analysis in Fundus Photography

Vitrectomy Eye Surgery - OPTIMAX

EZ Ancor Hollow Door downloadable

Imaging Unlimited