Artificial Intelligence in Cataract Surgery: A Systematic Review.

Purpose: The purpose of this study was to assess the current use and reliability of artificial intelligence (AI)-based algorithms for analyzing cataract surgery videos. Methods: A systematic review of the literature about intra-operative analysis of cataract surgery videos with machine learning techniques was performed. Cataract diagnosis and detection algorithms were excluded. Resulting algorithms were compared, descriptively analyzed, and metrics summarized or visually reported. The reproducibility and reliability of the methods and results were assessed using a modified version of the Medical Image Computing and Computer-Assisted (MICCAI) checklist. Results: Thirty-eight of the 550 screened studies were included, 20 addressed the challenge of instrument detection or tracking, 9 focused on phase discrimination, and 8 predicted skill and complications. Instrument detection achieves an area under the receiver operator characteristic curve (ROC AUC) between 0.976 and 0.998, instrument tracking an mAP between 0.685 and 0.929, phase recognition an ROC AUC between 0.773 and 0.990, and complications or surgical skill performs with an ROC AUC between 0.570 and 0.970. Conclusions: The studies showed a wide variation in quality and pose a challenge regarding replication due to a small number of public datasets (none for manual small incision cataract surgery) and seldom published source code. There is no standard for reported outcome metrics and validation of the models on external datasets is rare making comparisons difficult. The data suggests that tracking of instruments and phase detection work well but surgical skill and complication recognition remains a challenge for deep learning. Translational Relevance: This overview of cataract surgery analysis with AI models provides translational value for improving training of the clinician by identifying successes and challenges.

Inter-Rater Reliability of Proliferative Diabetic Retinopathy Assessment on Wide-Field OCT-Angiography and Fluorescein Angiography.

Purpose: To assess inter-rater reliability in the detection of proliferative diabetic retinopathy (PDR) changes using wide-field optical coherence tomography angiography (WF-OCTA) versus fluorescein angiography (FA). Methods: This retrospective, cross-sectional study included patients with severe nonproliferative and PDR. Images were acquired with 12 × 12 mm WF-OCTA and FA with a 55° lens. Images were cropped to represent the exact same field of view. Qualitative (detection of neovascularization at the disc [NVD] and elsewhere [NVE], enlarged foveal avascular zone [FAZ], vitreous hemorrhage [VH]) and quantitative analyses (FAZ area, horizontal, vertical, and maximum FAZ diameter) were performed by 2 masked graders using ImageJ. Inter-rater reliability was calculated using unweighted Cohen's kappa coefficient (κ) for qualitative analyses and intraclass correlation coefficients (ICC) for quantitative analyses. Results: Twenty-three eyes of 17 patients were included. Inter-rater reliability was higher for FA than for WF-OCTA in qualitative analyses: κ values were 0.65 and 0.78 for detection of extended FAZ, 0.83 and 1.0 for NVD, 0.78 and 1.0 for NVE, and 0.19 and 1 for VH for WF-OCTA and FA, respectively. In contrast, inter-rater reliability was higher for WF-OCTA than for FA in the quantitative analyses: ICC values were 0.94 and 0.76 for FAZ size, 0.92 and 0.79 for horizontal FAZ diameter, 0.82 and 0.72 for vertical FAZ diameter, and 0.88 and 0.82 for maximum FAZ diameter on WF-OCTA and FA, respectively. Conclusions: Inter-rater reliability of FA is superior to WF-OCTA for qualitative analyses whereas inter-rater reliability of WF-OCTA is superior to FA for quantitative analyses. Translational Relevance: The study highlights the specific merits of both imaging modalities in terms of reliability. FA should be preferred for qualitative parameters, whereas WF-OCTA should be preferred for quantitative parameters.