Automated Classification of Inherited Retinal Diseases in Optical Coherence Tomography Images Using Few-shot Learning.

Qi ZHAO; Si Wei MAI; Qian LI; Guan Chong HUANG; Ming Chen GAO; Wen Li YANG; Ge WANG; Ya MA; Lei LI; Xiao Yan PENG

Return

Automated Classification of Inherited Retinal Diseases in Optical Coherence Tomography Images Using Few-shot Learning.

Author: Qi ZHAO ¹ ; Si Wei MAI ² ; Qian LI ¹ ; Guan Chong HUANG ³ ; Ming Chen GAO ³ ; Wen Li YANG ¹ ; Ge WANG ¹ ; Ya MA ⁴ ; Lei LI ¹ ; Xiao Yan PENG ¹
Author Information

1. Department of Ophthalmology, Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University, Beijing Key Laboratory of Ophthalmology and Visual Sciences, Beijing 100730, China.
2. Department of Computer Science, Rutgers, The State University of New Jersey, New Brunswick 08901, USA.
3. Department of Computer Science and Engineering, University at Buffalo, Buffalo 14260, USA.
4. Beijing Institute of Ophthalmology, Beijing Tongren Hospital, Capital Medical University, Beijing Ophthalmology and Visual Science Key Laboratory, Beijing 100730, China.
Publication Type:Journal Article
Keywords: Few-shot learning; Inherited retinal diseases; Knowledge distillation; Optical coherence tomography; Retinal degeneration; Student-teacher learning; Transfer learning
MeSH: Humans; Tomography, Optical Coherence; Deep Learning; Retinal Diseases/diagnostic imaging*; Retina/diagnostic imaging*; ROC Curve
From: Biomedical and Environmental Sciences 2023;36(5):431-440
CountryChina
Language:English
Abstract: OBJECTIVE:To develop a few-shot learning (FSL) approach for classifying optical coherence tomography (OCT) images in patients with inherited retinal disorders (IRDs).
METHODS:In this study, an FSL model based on a student-teacher learning framework was designed to classify images. 2,317 images from 189 participants were included. Of these, 1,126 images revealed IRDs, 533 were normal samples, and 658 were control samples.
RESULTS:The FSL model achieved a total accuracy of 0.974-0.983, total sensitivity of 0.934-0.957, total specificity of 0.984-0.990, and total F1 score of 0.935-0.957, which were superior to the total accuracy of the baseline model of 0.943-0.954, total sensitivity of 0.866-0.886, total specificity of 0.962-0.971, and total F1 score of 0.859-0.885. The performance of most subclassifications also exhibited advantages. Moreover, the FSL model had a higher area under curves (AUC) of the receiver operating characteristic (ROC) curves in most subclassifications.
CONCLUSION:This study demonstrates the effective use of the FSL model for the classification of OCT images from patients with IRDs, normal, and control participants with a smaller volume of data. The general principle and similar network architectures can also be applied to other retinal diseases with a low prevalence.