Abstract: Detection of Optic disc (OD) in a fundus image is one of the important step in the process of automated screening of diabetic retinopathy. Hard Exudates detection algorithms generally find lot of false positives at the OD region since the intensity and colour distribution of OD region will much resemble that of a Hard Exudates region. So, most of the Exudates detection algorithms will wrongly classify the pixels at the OD region as Hard Exudates or Soft Exudates. In our previous work, we used Genetic Algorithm(GA) to find the OD location and size and reduced overall time, even doing the search on the entire problem space. In this work we are going to remove false hard exudates and will improving Retinopathy detection accuracy using GA based automatic optic disc detection and removal method. We are going to use the outputs of baseline hard exudates detection method presented in the of DIARETDB1[8] for evaluating the improvement after the removal of false positives in the OD region. We will evaluate the improvement in performance using the more versatile metric ‘weighted error rate’ (WER). After removing false positives at OD region by manual removal method and the proposed automatic removal method, we will compare the results of baseline method for evaluating the improvement in classification accuracy.