Automated morphological characterisation of the human knee joint for personalized arthroplasty

People

Supervisor

Description

This project is seeking to develop an automated pipeline for the extraction and processing of of morphological information of human knee joints from clinical images (X-ray and CT), with the ultimate purpose of generating coordinate sets for patient-specific virtual positioning of knee replacement implants. Secondary activities may include a population-level analysis of extracted morphological characteristics from a large clinical dataset.

An initial prototype pipeline was developed by previous students using Matlab. 2022 candidates will conduct a review of the existing solution and work on enhancing its methodological rigour, robustness of results and user-friendliness.

The project is of an applied and multi-disciplinary nature. Collaborative work with teams at the ANU School of Medicine and other interstate researchers will likely form part of the project activities.

Students applying for this project will ideally have good programming skills and prior experience with 2D and 3D image processing tasks. Good independent command of Computer Vision techniques would be highly beneficial. Prior biomedical knowledge or experience is very welvome but not required. We anticipate that the project would naturally attract Master of ML/CV students, though we will also gladly consider High-achieving undergraduate students with relevant interests and extracurricular experience.

Updated:  10 August 2021/Responsible Officer:  Dean, CECS/Page Contact:  CECS Marketing