Image analytics for medical imaging and forensic applications

Date
2016
Authors
Rajendran, Rahul
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract

Cancer is currently the second leading cause of death worldwide becoming a clear public health problem, both in developed and developing countries. Kidney cancer, also known as Renal cancer, is among the 10 most common cancers in both men and women. Image analysis plays a crucial role in the field of medicine, as the analysis guides the radiologist towards perfect diagnosis and treatment planning. At times, for faster acquisition, the medical imaging system that captures these tumors are of poor quality and contain a lot of distortions. Hence, enhancing an image to obtain better quality and preserving key features is an important issue in medical image processing. However, there does not exist a Computer Aided Classification (CAC) framework for the classification of renal cancer, and there is limited information about enhancing and segmenting kidney cancer images.

The goal of this thesis is to develop an image processing tool to enhance and segment medical and forensic images for the betterment of analysis. The key contributions of this thesis are four fold. First, we propose two enhancement techniques to enhance medical images for better analysis of renal cancer. We also propose a color correction technique to match touchless finger images. Second, we present a segmentation method to extract the tumor regions of the kidney. Third, Particle Swarm Optimization technique is used to extract the nuclei from histopathology images. This technique is further used in the elimination of background in fingerprint images. Finally, we analyze the histopathology images and generate an automated imaging system that classifies such images as benign and malignant.

Description
This item is available only to currently enrolled UTSA students, faculty or staff.
Keywords
Citation
Department
Electrical and Computer Engineering