Towards Ophthalmologist Level Accurate Deep Learning System for OCT Screening and Diagnosis
This addresses the need for automated, accurate OCT screening in ophthalmology, though it appears incremental as it builds on existing deep learning methods.
The researchers developed a fully automated deep learning system for analyzing retinal OCT images to screen and diagnose diseases, achieving state-of-the-art performance on the publicly available Mendeley OCT dataset.
In this work, we propose an advanced AI based grading system for OCT images. The proposed system is a very deep fully convolutional attentive classification network trained with end to end advanced transfer learning with online random augmentation. It uses quasi random augmentation that outputs confidence values for diseases prevalence during inference. Its a fully automated retinal OCT analysis AI system capable of pathological lesions understanding without any offline preprocessing/postprocessing step or manual feature extraction. We present a state of the art performance on the publicly available Mendeley OCT dataset.