Medical Image Restoration in MR CT Images of Human Abdomen

Abstract

Recent developments in automated analysis of medical images newlineusing advanced soft computing systems support healthcare professionals newlinein diagnostic tasks related with medical image interpretation. In medical newlineimaging applications, an accurate diagnosis and flawless assessment of newlinediseases greatly depends on factors like quality of image acquisition and newlineimage interpretation techniques. Advancements in Computer Aided newlineDesign (CAD) have enhanced the accuracy of interpretation of medical newlineimages from different sources like Computer Tomography (CT), newlineMagnetic Resonance (MR) and Ultrasound (US) imaging. However, for newlinebetter interpretation and diagnosis, medical images have to be free from newlinenoises, artifacts, etc. The noise components in the captured images newlinedeteriorate the image quality in applications even amidst continuous newlineimprovement in the fields of medical imaging, image acquisition and newlineprocessing. Thus, removing these noise components from medical newlineimages is a major challenge in medical image signal processing domain. newlineThe aim of this research is developing an efficient denoising newlineand image restoration technique for medical images from different newlinesources like CT, MRI and US scan images of Human abdomen. newlineDifferent algorithms like Hybrid Filter based Firefly Algorithm, Radial newlineix newlineBasis Function Neural Network, Cuckoo Search Optimization newlineTechnique with Extended Kalman filter and Grey Wolf Optimization newline(GWO) with Extended Kalman filter (EKF) technique were developed newlinein this work. newlineMATLAB 15a software environment was used for newlineimplementing the proposed image denoising and restoration algorithms. newlineThe dataset used for evaluating and analysing the algorithms were newlinederived from public dataset consisting of medical images of CT, MR newlineand US images of Human Abdomen including Kidney, Pancreas and newlineLiver. The performance of the proposed techniques were assessed using newlinedifferent evaluation parameters. The results of evaluation show that newlineGWO EKF method for the image restoration and denoising has newlinegenerated better results across different medical ima