Computer Aided Diagnosis of Idiopathic Central Serous Chorioretinopathy

Abstract

C244 Aided Diagnosis (CAD) using Artificial Intelligence (AI) techniques is being widely used these days especially in the field of medical imaging. Central Serous Chorioretinopathy (CSCR) is a common visual impairment syndrome that is predominately found amongst the young age fellows. Its main reason is the abnormal accumulation of serous in macula which is the central region of retina layer. Optical Coherence Tomography (OCT) is a non-invasive mean of eye inspection that helps in the early diagnosis of retina related diseases more efficiently than traditional methods. In this paper we report a CAD based fully automated algorithm for the detection of CSCR affected subjects using k-NN classifier. We extracted three discrete features based on the thickness of the retina layer and area of serous deposits. We have studied 80 retinal subjects (40 healthy and 40 CSCR affected OCT scans) during our research and accurately classified all the test cases for a value of k = 5, achieving 100% efficiency.