Glaucoma detection: Binocular approach and clinical data in machine learning

Abstract

In this work, we present a multi-modal machine learning method to automate early glaucoma diagnosis. The proposed methodology introduces two novel aspects for automated diagnosis not previously explored in the literature: simultaneous use of ocular fundus images from both eyes and integration with the patient’s additional clinical data. We begin by establishing a baseline, termed monocular mode , which adheres to the traditional approach of considering the data from each eye as a separate instance. We then explore the binocular mode , investigating how combining information from both eyes of the same patient can enhance glaucoma diagnosis accuracy. This exploration employs the PAPILA dataset, comprising information from both eyes, clinical data, ocular fundus images, and expert segmentation of these images. Additionally, we compare two image-derived data modalities: direct ocular fundus images and morphological data from manual expert segmentation. Our method integrates Gradient-Boosted Decision Trees (GBDT) and Convolutional Neural Networks (CNN), specifically focusing on the MobileNet, VGG16, ResNet-50, and Inception models. SHAP values are used to interpret GBDT models, while the Deep Explainer method is applied in conjunction with SHAP to analyze the outputs of convolutional-based models. Our findings show the viability of considering both eyes, which improves the model performance. The binocular approach, incorporating information from morphological and clinical data yielded an AUC of 0.796 ( ± 0 . 003 at a 95% confidence interval), while the CNN, using the same approach (both eyes), achieved an AUC of 0.764 ( ± 0 . 005 at a 95% confidence interval). • To our knowledge, this is the first study to use information from both eyes of the same patient to automatically diagnose glaucoma. • We provide a comparative analysis highlighting the improvements in diagnosis when utilizing bilateral eye information versus an eye-by-eye approach. • The research demonstrates the efficacy of using data from both eyes to enhance early-stage glaucoma diagnosis in both gradient-boosted decision trees (GBDT) and convolutional neural networks (CNN) models. • Additionally, we investigate the impact of incorporating patients’ clinical test results on the accuracy of glaucoma diagnosis. • Furthermore, we interpret model insights using SHAP values and the Deep Explainer method for CNN-based models. • The proposed model is interpretable, allowing analysis of each input’s contribution to the final diagnosis. • With the GBDT models that combine data from both eyes and integrate clinical and morphological features from ocular fundus images, we achieved an AUC of 0.796 ± 0.003 at a 95% confidence interval. In the CNN-based approach, we attained 0.764 ± 0.005 at the same confidence level.

Topics & Concepts