Advancements in Medical Image Processing Techniques: From Computerized Diagnosis to Blockchain

Abstract

Healthcare providers and researchers are advancing their image-guided interventions and health management system to provide a flexible, efficient, and early diagnosis for patients. Imaging informatics aims to improve the efficiency, accuracy, usability, and reliability of medical imaging services within the healthcare enterprise. In particular, medical imaging is an essential tool used to create a visual representation of a human body for clinical analysis and medical interventions. This processing includes many types of techniques and operations such as image gaining, analysis, artifact elimination, intelligent noise reduction, storage, and communication. Therefore, it’s important to focus on research related to medical imaging technologies for routine clinical practice that helps doctors to access and interpret medical images accurately. This dissertation focuses on solving three distinct issues related to medical imaging techniques, i.e., 1) image analysis 2) image reconstruction and 3) image management. This thesis aims to analyze and enhance medical image processing with application to diagnostic ultrasound. Furthermore, improving the medical image storage and sharing system using decentralized architecture. Our advanced solutions based on computer-aided diagnosis, computational imaging, and blockchain. In this thesis, we firstly categorize and review the computer-aided diagnostic (CAD) system according to the four primary stages including data pre-processing, lesion segmentation, feature extraction, selection, and classifier. The purpose of a CAD system is to process and provide adequate information that helps to analyze the ultrasound images for early cancer diagnosis. Secondly, we focused on the image preprocessing stage for reducing multiplicative noise present on ultrasound image using a sparse representation over a learned dictionary. Finally, we propose a proof-of-concept (POC) design for a distributed framework called a patient-centric image management (PCIM) system that is a blockchain-based architecture designed to facilitate secured patient-centric access and storage of encrypted medical images within an open distributed network.