Comparison of Machine and Human Expert Evaluation of Capsulorrhexis Creation Performance Through Analysis of Surgical Video Recordings.

Purpose: Achieving competency in cataract surgery is an essential component of ophthalmology residency training. Video-based analysis of surgery can change training through its objective, reliable, and timely assessment of resident performance. Methods: Using the Image Labeler application in MATLAB, the capsulorrhexis step of 208 surgical videos, recorded at the University of Michigan, was annotated for subjective and objective analysis. Two expert surgeons graded the creation of the capsulorrhexis based on the International Council of Ophthalmology's Ophthalmology Surgical Competency Assessment Rubric:Phacoemulsification (ICO-OSCAR:phaco) rating scale and a custom rubric (eccentricity, roundness, size, centration) that focuses on the objective aspects of this step. The annotated rhexis frames were run through an automated analysis to obtain objective scores for these components. The subjective scores were compared using both intra and inter-rater analyses to assess the consistency of a human-graded scale. The subjective and objective scores were compared using intraclass correlation methods to determine relative agreement. Results: All rhexes were graded as 4/5 or 5/5 by both raters for both items 4 and 5 of the ICO-OSCAR:phaco rating scale. Only roundness scores were statistically different between the subjective graders (mean difference = -0.149, p-value = 0.0023). Subjective scores were highly correlated for all components (>0.6). Correlations between objective and subjective scores were low (0.09 to 0.39). Conclusion: Video-based analysis of cataract surgery presents significant opportunities, including the ability to asynchronously evaluate performance and provide longitudinal assessment. Subjective scoring between two raters was moderately correlated for each component.

Evaluation of Methods for Detection and Semantic Segmentation of the Anterior Capsulotomy in Cataract Surgery Video.

Background: The capsulorhexis is one of the most important and challenging maneuvers in cataract surgery. Automated analysis of the anterior capsulotomy could aid surgical training through the provision of objective feedback and guidance to trainees. Purpose: To develop and evaluate a deep learning-based system for the automated identification and semantic segmentation of the anterior capsulotomy in cataract surgery video. Methods: In this study, we established a BigCat-Capsulotomy dataset comprising 1556 video frames extracted from 190 recorded cataract surgery videos for developing and validating the capsulotomy recognition system. The proposed system involves three primary stages: video preprocessing, capsulotomy video frame classification, and capsulotomy segmentation. To thoroughly evaluate its efficacy, we examined the performance of a total of eight deep learning-based classification models and eleven segmentation models, assessing both accuracy and time consumption. Furthermore, we delved into the factors influencing system performance by deploying it across various surgical phases. Results: The ResNet-152 model employed in the classification step of the proposed capsulotomy recognition system attained strong performance with an overall Dice coefficient of 92.21%. Similarly, the UNet model with the DenseNet-169 backbone emerged as the most effective segmentation model among those investigated, achieving an overall Dice coefficient of 92.12%. Moreover, the time consumption of the system was low at 103.37 milliseconds per frame, facilitating its application in real-time scenarios. Phase-wise analysis indicated that the Phacoemulsification phase (nuclear disassembly) was the most challenging to segment (Dice coefficient of 86.02%). Conclusion: The experimental results showed that the proposed system is highly effective in intraoperative capsulotomy recognition during cataract surgery and demonstrates both high accuracy and real-time capabilities. This system holds significant potential for applications in surgical performance analysis, education, and intraoperative guidance systems.

Development of Crystalline Corneal Opacities (Steroid Keratopathy) in Dogs After Treatment With Ophthalmic Corticosteroids.

PURPOSE: To retrospectively evaluate and describe the relationship between the use of topical corticosteroids and the development of crystalline corneal opacities (steroid keratopathy) in a colony of research Beagles and Beagle-derived dogs. METHODS: Medical records of 73 purpose-bred Beagles and Beagle-derived dogs were reviewed from June 2012 to May 2021. All dogs were treated with topical ophthalmic corticosteroids for at least 21 days. In addition to regular ophthalmic examination, some dogs also had a systemic lipid profile (n = 6) performed to work up further and characterize the crystalline corneal opacities. Globes of 3 dogs were examined histopathologically. RESULTS: Axial stromal crystalline corneal opacities were appreciated in 25 eyes of 14 dogs after a median of 141 days after initiating treatment (35-396 days). Multiple corticosteroids were used, including neomycin-polymyxin b-dexamethasone 0.1% ophthalmic ointment, prednisolone acetate 1% ophthalmic suspension, and difluprednate 0.05% ophthalmic emulsion (Durezol). Resolution of corneal opacity was documented in 4 of 25 eyes when ophthalmic corticosteroids were discontinued after a median of 406.5 days (271-416 days). Histopathologic examination revealed a dense band of acellular material, poorly staining with periodic acid-Schiff, subtending the corneal epithelium, and being surrounded by spindle cells. CONCLUSIONS: This case series documents the onset of steroid keratopathy in Beagles and Beagle-derived dogs after treatment with ophthalmic corticosteroids. Clinical resolution of steroid keratopathy lesions may be possible after discontinuation of ophthalmic corticosteroids.

Severe Spontaneous Tilt of Scleral-Fixated Intraocular Lenses.

PURPOSE: To report and evaluate a multicenter series of 18 cases of severe, spontaneous IOL tilt involving the flanged intrascleral haptic fixation technique (FISHF). DESIGN: Clinical study with historical controls. METHODS: We report a cross-sectional study of 46 FISHF cases using the CT Lucia 602 IOL at a single academic center over a period of 24 weeks to determine the incidence of severe rotisserie-style rotational tilt. These rates were then compared with the same time-frame the prior year to help determine if this is a new phenomenon. Additional cases of severe tilt were solicited from another 4 academic centers. RESULTS: Among 46 FISHF cases at a single center, 5 developed severe tilt. No clear pattern in surgical technique, ocular history, or ocular anatomy was evident in these cases compared with controls, although the involved IOLs clustered within a narrow diopter range, indicative of a batch effect. In the same 24-week interval the year before, 33 FISHF cases were performed, none of which exhibited severe rotational tilt. In our multicenter dataset, 18 cases of tilt were identified. Surgeons included fellow and early-career physicians as well as surgeons with multiple years of experience with the Yamane technique. A variety of surgical approaches for FISHF were represented. In at least 8 of the cases, haptic rotation and/or dehiscence at the optic-haptic junction were documented. CONCLUSIONS: The identification of haptic rotation and dehiscence intraoperatively in several cases may reflect a new stability issue involving the optic-haptic junction.

Adaptive Tensor-Based Feature Extraction for Pupil Segmentation in Cataract Surgery.

Cataract surgery remains the only definitive treatment for visually significant cataracts, which are a major cause of preventable blindness worldwide. Successful performance of cataract surgery relies on stable dilation of the pupil. Automated pupil segmentation from surgical videos can assist surgeons in detecting risk factors for pupillary instability prior to the development of surgical complications. However, surgical illumination variations, surgical instrument obstruction, and lens material hydration during cataract surgery can limit pupil segmentation accuracy. To address these problems, we propose a novel method named adaptive wavelet tensor feature extraction (AWTFE). AWTFE is designed to enhance the accuracy of deep learning-powered pupil recognition systems. First, we represent the correlations among spatial information, color channels, and wavelet subbands by constructing a third-order tensor. We then utilize higher-order singular value decomposition to eliminate redundant information adaptively and estimate pupil feature information. We evaluated the proposed method by conducting experiments with state-of-the-art deep learning segmentation models on our BigCat dataset consisting of 5,700 annotated intraoperative images from 190 cataract surgeries and a public CaDIS dataset. The experimental results reveal that the AWTFE method effectively identifies features relevant to the pupil region and improved the overall performance of segmentation models by up to 2.26% (BigCat) and 3.31% (CaDIS). Incorporation of the AWTFE method led to statistically significant improvements in segmentation performance (P < 1.29 × 10-10 for each model) and yielded the highest-performing model overall (Dice coefficients of 94.74% and 96.71% for the BigCat and CaDIS datasets, respectively). In performance comparisons, the AWTFE consistently outperformed other feature extraction methods in enhancing model performance. In addition, the proposed AWTFE method significantly improved pupil recognition performance by up to 2.87% in particularly challenging phases of cataract surgery.

CatStep: Automated Cataract Surgical Phase Classification and Boundary Segmentation Leveraging Inflated 3D-Convolutional Neural Network Architectures and BigCat.

Objective: Accurate identification of surgical phases during cataract surgery is essential for improving surgical feedback and performance analysis. Time spent in each surgical phase is an indicator of performance, and segmenting out specific phases for further analysis can simplify providing both qualitative and quantitative feedback on surgical maneuvers. Study Design: Retrospective surgical video analysis. Subjects: One hundred ninety cataract surgical videos from the BigCat dataset (comprising nearly 4 million frames, each labeled with 1 of 11 nonoverlapping surgical phases). Methods: Four machine learning architectures were developed for segmentation of surgical phases. Models were trained using cataract surgical videos from the BigCat dataset. Main Outcome Measures: Models were evaluated using metrics applied to frame-by-frame output and, uniquely in this work, metrics applied to phase output. Results: The final model, CatStep, a combination of a temporally sensitive model (Inflated 3D Densenet) and a spatially sensitive model (Densenet169), achieved an F1-score of 0.91 and area under the receiver operating characteristic curve of 0.95. Phase-level metrics showed considerable boundary segmentation performance with a median absolute error of phase start and end time of just 0.3 seconds and 0.1 seconds, respectively, a segmental F1-score @70 of 0.94, an oversegmentation score of 0.89, and a segmental edit score of 0.92. Conclusion: This study demonstrates the feasibility of high-performance automated surgical phase identification for cataract surgery and highlights the potential for improved surgical feedback and performance analysis. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.