Adam Optimized Deep Learning Model for Segmenting ROI Region in Medical Imaging

Abstract

Medical image processing played a crucial role in healthcare sectors because it produces visible images of inner body structure and internal tissues. With medical imaging, the organs' functions are analyzed, and anomalies are detected effectively. The imaging process utilizes enhancement, analysis, and recognition procedures that maximize disease detection accuracy. However, the gathered medical images have noise while performing the medical research that affects the ROI region segmentation process. The improper identification of the ROI region causes to reduce disease detection accuracy and create complexity. Therefore, this paper uses the Adam optimization technique with deep learning approaches for examining the medical images. The deep learning (DL) approaches utilizing the multiple layers, expert-tuned parameters, and learning function to deriving the affected ROT region. The DL method automatically extracts the medical imaging features that are optimized according to the RMS prop and momentum-based learning parameters. The optimized parameter minimizes the complexity while deriving the disease affected region. This process ensures high region segmentation accuracy with minimum time. The process has detected the affected region with 99.27% of accuracy.