Biomedical Imaging

Background

Medical imaging techniques have obtained great development in recent decades and have been found within many different applications in healthcare. Medical doctors use medical imaging technologies such as ultrasound and X-ray imaging to diagnosis diseases. Scientists use modern imaging technologies to understand the functions of the brain and the heart. The research of medical imaging at Michigan Technological University focuses on the development of new imaging technologies, such as 3-D ultrasound imaging, OCT (Optical coherence tomography) imaging, and optical imaging, and the development of new image analysis algorithms to support disease detection and diagnosis.


 

Research Project I

Title: Computer aided skin cancer detection and diagnosis

Faculty Investigator: Jinshan Tang

email: jinshant at mtu dot edu

Summary: Cancer is one of the biggest threats to human beings and skin cancer is the most prevalent form of cancers in USA. But if skin cancer can be diagnosed in its early phase, the curable rate will be more than 90%. So the aim of this project is to develop a self-diagnosis tool to help people to detect skin cancer at its early stage. In this project, we mainly study the key techniques used in computer aided cancer detection system (CADs) for skin cancer detection, such as the extraction of the boundary of skin lesions.  Besides, we also focus on the development of an iPhone application which can be used for skin cancer detection by persons which are not dermatologists. The iPhone based skin cancer detection application is a care-everywhere tool which provides people with a good skin care method and can help a person find skin cancer without going to a clinic.  

 

Research Project II

Title: Computer aided breast cancer detection and diagnosis using mammography

Faculty Investigator: Jinshan Tang

email: jinshant at mtu dot edu

Summary: The aim of this project is to develop effective techniques which are needed to improve the performance of the current computer aided breast cancer detection systems. These techniques include image processing and pattern recognition techniques for the detection of calcifications, masses, architectural distortion, bilateral asymmetry, and machine learning techniques for automatic classification of benign and malign areas in breasts.

 

Research Project III

Title: GPU accelerated 3-D ultrasound image processing technologies

Faculty Investigator: Jinshan Tang

email: jinshant at mtu dot edu

Summary: 3-D ultrasound technology has been proven to exceed 2-D ultrasound technology in monitoring individual follicles, determining the patterns of follicular development, predicting early pregnancy, determining fetal sex, and diagnosing the abnormalities of the reproductive organs. However, because of the large size of the 3-D data, the computation in the applications of 3-D ultrasound technology is time consuming. Thus, we are developing new technologies such as GPU technology to accelerate the computation in the applications of 3-D ultrasound technology. The current focus of the project is to develop GPU technology for speckle reduction and object segmentation in 3-D ultrasound images.