Exploiting the Transferability of Deep Learning Systems Across Multi-modal Retinal Scans for Extracting Retinopathy Lesions

TL;DR

This study evaluates deep learning models for retinal lesion segmentation across multi-modal scans and introduces a transferability strategy, demonstrating high accuracy and generalization across diverse datasets.

Contribution

It presents a novel transferability approach for deep learning models in retinal lesion detection across different scanner types and imaging modalities.

Findings

Hybrid RAGNet achieved a mean dice score of 0.822.

Models generalized well across multiple datasets.

Source code is publicly available.

Abstract

Retinal lesions play a vital role in the accurate classification of retinal abnormalities. Many researchers have proposed deep lesion-aware screening systems that analyze and grade the progression of retinopathy. However, to the best of our knowledge, no literature exploits the tendency of these systems to generalize across multiple scanner specifications and multi-modal imagery. Towards this end, this paper presents a detailed evaluation of semantic segmentation, scene parsing and hybrid deep learning systems for extracting the retinal lesions such as intra-retinal fluid, sub-retinal fluid, hard exudates, drusen, and other chorioretinal anomalies from fused fundus and optical coherence tomography (OCT) imagery. Furthermore, we present a novel strategy exploiting the transferability of these models across multiple retinal scanner specifications. A total of 363 fundus and 173,915 OCT scans from seven publicly available datasets were used in this research (from which 297 fundus and 59,593 OCT scans were used for testing purposes). Overall, a hybrid retinal analysis and grading network (RAGNet), backboned through ResNet-50, stood first for extracting the retinal lesions, achieving a mean dice coefficient score of 0.822. Moreover, the complete source code and its documentation are released at: http://biomisa.org/index.php/downloads/.