An Advanced Lcl-Dlnn Algorithm Based Retinal Disease Detection Using Retinal Fundus Image

Abstract

Medical image processing has recently increased the precision with which medical images are used to
identify disorders in people. Using Deep Learning (DL) methodologies, which allow automatic learning of
the associated characteristics for specific tasks rather than handmade procedures, has significantly
improved the analysis of medical images. Currently, a variety of methods for autonomously segmenting
retinal fundus images have been developed. However, to increase computational complexity and
decreased efficiency, they were unable to provide superior accuracy. In this study, Learning Curved
Layered Deep Learning Neural Network (LCL-DLNN) is used to suggest an effective detection method
employing retinal fundus images. The retinal images are first pre-processed in two processes, such as
image cropping and image contrast level, to improve the image quality. Second, to increase the input
image quality from segmented images and increase prediction accuracy, we design the Edge Stop
Functional Iterative Region Growing (ESFIRG) method. Finally, the retinal fundus image can be used to
forecast whether or not the proposed LCL-DLNN approach would perform. The suggested method is
implemented on the MATLAB platform, and performance is assessed using presentation metrics. The
suggested technique outperforms the currently used research approaches in experimental analysis.