Unsupervised corneal contour extraction algorithm with shared model for dynamic deformation videos: improving accuracy and noise resistance.

BACKGROUND: In this study, an automatic corneal contour extraction algorithm with a shared model is developed to extract contours from dynamic corneal videos containing noise, which improves the accuracy of corneal biomechanical evaluation and clinical diagnoses. The algorithm does not require manual labeling and completes the unsupervised semantic segmentation of each frame in corneal dynamic deformation videos based on a fully convolutional deep-learning network using corneal geometry and texture information. RESULTS: We included 1027 corneal videos at Tianjin Eye Hospital (Nankai University Affiliated Eye Hospital) from May 2020 to November 2021. The videos were obtained by the ultra-high-speed Scheimpflug camera, and then we used the shared model mechanism to accelerate the segmentation of corneal regions in videos, effectively resist noise, determine corneal regions based on shape factors, and finally achieve automatic and accurate extraction of corneal region contours. The Intersection over Union (IoU) of the extracted and real corneal contours using this algorithm reached 95%, and the average overlap error was 0.05, implying that the extracted corneal contour overlapped almost completely with the real contour. CONCLUSIONS: Compared to other algorithms, the method introduced in this study does not require manual annotation of corneal contour data in advance and can still extract accurate corneal contours from noisy corneal videos with good repeatability.

Localized Corneal Biomechanical Alteration Detected In Early Keratoconus Based on Corneal Deformation Using Artificial Intelligence.

PURPOSE: This study aimed to develop a novel method to diagnose early keratoconus by detecting localized corneal biomechanical changes based on dynamic deformation videos using machine learning. DESIGN: Diagnostic research study. METHODS: We included 917 corneal videos from the Tianjin Eye Hospital (Tianjin, China) and Shanxi Eye Hospital (Xi'an, China) from February 6, 2015, to August 25, 2022. Scheimpflug technology was used to obtain dynamic deformation videos under forced puffs of air. Fourteen new pixel-level biomechanical parameters were calculated based on a spline curve equation fitting by 115,200-pixel points from the corneal contour extracted from videos to characterize localized biomechanics. An ensemble learning model was developed, external validation was performed, and the diagnostic performance was compared with that of existing clinical diagnostic indices. The performance of the developed machine learning model was evaluated using precision, recall, F1 score, and the area under the receiver operating characteristic curve. RESULTS: The ensemble learning model successfully diagnosed early keratoconus (area under the curve = 0.9997) with 95.73% precision, 95.61% recall, and 95.50% F1 score in the sample set (n=802). External validation on an independent dataset (n=115) achieved 91.38% precision, 92.11% recall, and 91.18% F1 score. Diagnostic accuracy was significantly better than that of existing clinical diagnostic indices (from 86.28% to 93.36%, all P <0.01). CONCLUSIONS: Localized corneal biomechanical changes detected using dynamic deformation videos combined with machine learning algorithms were useful for diagnosing early keratoconus. Focusing on localized biomechanical changes may guide ophthalmologists, aiding the timely diagnosis of early keratoconus and benefiting the patient's vision.

Artificial Intelligence-Based Diagnostic Model for Detecting Keratoconus Using Videos of Corneal Force Deformation.

Purpose: To develop a novel method based on biomechanical parameters calculated from raw corneal dynamic deformation videos to quickly and accurately diagnose keratoconus using machine learning. Methods: The keratoconus group was included according to Rabinowitz's criteria, and the normal group included corneal refractive surgery candidates. Independent biomechanical parameters were calculated from dynamic corneal deformation videos. A novel neural network model was trained to diagnose keratoconus. Tenfold cross-validation was performed, and the sample set was divided into a training set for training, a validation set for parameter validation, and a testing set for performance evaluation. External validation was performed to evaluate the model's generalizability. Results: A novel intelligent diagnostic model for keratoconus based on a five-layer feedforward network was constructed by calculating four biomechanical characteristics, including time of the first applanation, deformation amplitude at the highest concavity, central corneal thickness, and radius at the highest concavity. The model was able to diagnose keratoconus with 99.6% accuracy, 99.3% sensitivity, 100% specificity, and 100% precision in the sample set (n = 276), and it achieved an accuracy of 98.7%, sensitivity of 97.4%, specificity of 100%, and precision of 100% in the external validation set (n = 78). Conclusions: In the absence of corneal topographic examination, rapid and accurate diagnosis of keratoconus is possible with the aid of machine learning. Our study provides a new potential approach and sheds light on the diagnosis of keratoconus from a purely corneal biomechanical perspective. Translational Relevance: Our findings could help improve the diagnosis of keratoconus based on corneal biomechanical properties.

[GSH fermentation process modeling using entropy-criterion based RBF neural network model].

The prediction accuracy and generalization of GSH fermentation process modeling are often deteriorated by noise existing in the corresponding experimental data. In order to avoid this problem, we present a novel RBF neural network modeling approach based on entropy criterion. It considers the whole distribution structure of the training data set in the parameter learning process compared with the traditional MSE-criterion based parameter learning, and thus effectively avoids the weak generalization and over-learning. Then the proposed approach is applied to the GSH fermentation process modeling. Our results demonstrate that this proposed method has better prediction accuracy, generalization and robustness such that it offers a potential application merit for the GSH fermentation process modeling.