A semi-automatic computer aided system for spine diagnosis

This project aims at developing a computerized system for spine diagnosis. This system will improve the efficiency and efficacy of radiologists to diagnose patients’ spine problems. The development of the system involves devising a set of tailor-made mathematical formulation. These formulation are grounded on the state-of-the-art computer vision algorithms and they are capable of capturing the knowledge required during spine diagnosis. The computerized system employing these mathematical formulations will be able to mimic the human expert to perform basic image-based diagnosis of spine. The system facilitates standardizing the diagnostic procedures and criterion in order to suppress interior intra-observer variability and subsequently benefit disease progress tracking. The spinal image statistics captured by the system will guide radiologists to focus on the problematic regions and as a result, radiologists’ time required for diagnosis will be reduced. These allow diagnosis of more cases based on the same working hours. Processing extra cases also leads to additional incomes to the professionals in the partner organization.

Wai Kong Law
Faculty Supervisor: 
Dr. Andrew Leung