Gagan Kalra

Machine Learning-Based Automated Detection and Quantification of Geographic Atrophy and Hypertransmission Defects Using Spectral Domain Optical Coherence Tomography

, , , , , , , , , , , , , ,

The current study describes the development and assessment of innovative, machine learning (ML)-based approaches for automated detection and pixel-accurate measurements of regions with geographic atrophy (GA) in late-stage age-related macular degeneration (AMD) using optical coherence tomography systems. 900 OCT volumes, 100266 B-scans, and en face OCT images from 341 non-exudative AMD patients with or without GA were included in this study from both Cirrus (Zeiss) and Spectralis (Heidelberg) OCT systems. B-scan and en face level ground truth GA masks were created on OCT B-scan where the segmented ellipsoid zone (EZ) line, retinal pigment epithelium (RPE) line, and bruchs membrane (BM) line overlapped. Two deep learning-based approaches, B-scan level and en face level, were trained. The OCT B-scan model had detection accuracy of 91% and GA area measurement accuracy of 94%. The en face OCT model had detection accuracy of 82% and GA area measurement accuracy of 96% with primary target...

Automated Machine Learning Detection of Hydroxychloroquine Toxicity and Prediction of Future Toxicity Using Higher Order OCT Biomarkers

, , , , , ,

Objective: Despite guidelines for HCQ toxicity screening, there are clear challenges to accurate detecting and interpretation. In the current report, the feasibility of automated machine-learning (ML) -based detection of hydroxychloroquine (HCQ) retinopathy and prediction of progression to toxicity in eyes without preexisting toxicity is described. Design: Retrospective longitudinal cohort study. Subjects and methods: This was an IRB-approved retrospective longitudinal image analysis study of 388 subjects on hydroxychloroquine. Multi-layer compartmental retinal segmentation with EZ mapping was utilized to harvest quantitative SDOCT biomarkers. Utilizing a combination of clinical features (i.e., cumulative HCQ dose, duration of therapy) and quantitative imaging biomarkers (e.g., volumetric ellipsoid zone (EZ) integrity and compartmental measurements), machine learning models were created to detect toxicity and predict progression based on ground truth OCT-based toxicity reads by 2 ma...

OCTAngiolytics: A Review of Quantitative OCT Angiography Biomarkers

, , , ,

With an increase in scan acquisition speeds, increased field of view and high-quality evidence proving unequivocal correlation with the established gold standard ultrawide-field fluorescein angiography, a lot of research has been conducted using optical coherence tomography angiography (OCTA) to generate quantitative vascular metrics for normal eyes and eyes with several retinal and choroidal diseases. With numerous biomarkers, methods of analysis, automated or manual or semiautomated approaches and a plethora of poorly interchangeable device-specific data described in literature, it is easy for clinicians to get overwhelmed when trying to get a holistic grasp of this topic. This video summarizes various described OCTA biomarkers, methods to achieve them and merits and challenges that come with them for analysis of chorioretinal vasculature in healthy eyes and various conditions. ( Read Full Article )

Semi-automated quantitative analysis of the middle limiting membrane in tubercular serpiginous-like choroiditis using swept-source optical coherence tomography

, , , ,

o analyze the longitudinal changes in the outer plexiform layer (OPL) in patients with tubercular serpiginous-like choroiditis (TB SLC) and compare it to the healthy control population. Clinical and imaging data of subjects with TB SLC (minimum 6-month follow-up) and healthy control subjects were reviewed. Optical coherence tomography (OCT) imaging obtained using swept-source device (DRI Triton, Topcon, Japan) from three visits (baseline, 3 months, and 6 months) were analyzed. Three OCT scans were chosen-one passing through the center of the fovea, one line above, and one line below. After random indexing to anonymize the images, they were pre-processed and fed into an automated pipeline to identify, crop, and measure the area of the OPL in the line scan. Longitudinal comparisons of OPL within the patient group were performed. The study included 32 eyes (16 patients; 11 males; mean age: 32.9 ± 7.8 years) with TB SLC. Twenty-eight eyes (14 subjects; 10 males: mean age: 31.1 &p...

Wide-field optical coherence tomography imaging in diabetic retinopathy

, , , , , , , ,

Purpose: To report the individual retinal layer thicknesses up to mid-equator in patients with diabetic retinopathy (DR) using Spectralis (Heidelberg Engineering, Heidelberg, Germany) wide-field optical coherence tomography (OCT). Methods: Retinal layers were segmented using a custom designed semi-automated algorithm, where reference points were marked by the examiner to enable software to automatically compute the thickness values of each retinal sublayer at an interval of 1 mm from reference points. The values of individual retinal thicknesses in eyes with varying severity of DR were compared with the values of healthy subjects. Generalized estimating equation was performed to compensate for inclusion of both eyes of patients. Results: A total of 64 patients (119 eyes) with a mean age of 68.97 ± 10.27 years were included. Overall, ganglion cell layer (GCL)/ inner plexiform layer (IPL) complex (-31.67 microns, p < 0.001), outer plexiform layer (-6.78 microns, p = 0.002) a...

Optical computed tomography (OCT) angiolytics: A review of OCT angiography quantiative biomarkers

, , , ,

Optical coherence tomography angiography (OCTA) provides a noninvasive method to obtain angiography of the chorioretinal vasculature leading to its recent widespread adoption. With growing number of studies exploring the use of OCTA, various biomarkers have been described to quantify vascular characteristics using various techniques. We summarize biomarkers currently described for retinal and choroidal vasculature OCTA systems and the methods used to obtain them. We presents a critical review of these methods and key findings in common retinal diseases andalso appraise future directions, including application of artificial intelligence in OCTA angiolytics. ( Read Full Article )

Recent advances in wide field and ultra-wide field optical coherence tomography angiography in retinochoroidal pathologies

, , , , , ,

Introduction : With the advent of wide field (WF) and ultra-wide field (UWF) imaging in color fundus photography and fluorescein angiography (FA), the importance of an expanded field-of-view (FOV) has been illustrated for early detection of several chorioretinal disorders. A need to bring similar utility to optical coherence tomography angiography (OCTA) is imperative. Areas covered : For this review, we explored all research items with WF or UWF on Pubmed Central and Google Scholar. Software based methods included algorithms for motion correction, artifact reduction, image pre-processing, post-processing and montage protocols while hardware methods included swept-source OCTA, faster scanning sources, hardware-based motion tracking and image registration, scanning sources with faster speeds and lower resolution. For various disorders such as diabetic retinopathy, uveitis, vascular disorders, among others, the increased FOV showed demonstrably increased detection rates and diagnostic...


Organizations in the News

OCT Companies in the News

Picture Gallery