Field: Biomedical engineering

Location: St Lucia campus

Type of student: Volunteer/extra-curricular

Brief synopsis: Magnetic resonance (MR) imaging, while an important non-invasive radiological modality for various clinical applications, are difficult to analyse due to the presence of a large number of complex structures. Thus, extracting meaningful clinical information without human interaction is a challenging task. The shape of bones and organs is one such particular type of clinical information that can be important for diagnosis and treatment, which can take hours of manual interpretation by a radiologist or clinician. In this project, the student will apply machine learning techniques to identify keypoints on bone or organ surfaces based on curvature of surface data already available from previous projects involving the supervisor. The learning the key curvature components of these types of surfaces, it is hoped that the resulting keypoints on the surface will lead to simplifying problems such as surface registration and shape modelling. This work will have applications to analysing human joints and cartilage from MR images and shape analysis of human organs that may eventually be available in MRI scanners.

Note: This is project is only suitable for students who have some machine learning and/or physics experience.

Prerequisite skills: Python/C/C++, Machine Learning, PDEs, Physics


Dr Shekhar Chandra

Dr Shekhar Chandra

School of Information Technology and Electrical Engineering