Investigation of Artificial Intelligence (AI) techniques to identify the relationship between neuroimaging and clinical assessments in Friedreich Ataxia

Friedreich Ataxia (FA) is a rare, inherited, progressive neurodegenerative disorder that yields unsteady, awkward movements, impaired sensory functions and heart disease/diabetes. Although rare, FA affects one in 50,000 people in the United States and is the most common inherited ataxia. Currently drug development for rare, neurodegenerative diseases like FA is challenging and limited by a lack of validated, sensitive biomarkers of disease progression, which may in turn facilitate targeted therapeutic interventions and evaluating their efficacy. Monash University is leading TRACK-FA, the first multi-site natural history study of neurobiological changes underlying FA and a world-first intensive effort to combine global expertise in FA. TRACK-FA is a collaboration between seven international universities and is sponsored by the Friedreich’s Ataxia Research Alliance (FARA) in the USA, as well as leading industry partners. The study aims to improve understanding of the natural history of FA, validate neuroimaging markers in FA to deliver a set of trial-ready biomarkers, and to develop a comprehensive database to facilitate ongoing community research and discovery. This PhD project involves the application of state-of-the-art machine learning and deep learning techniques on the pre-existing TRACK-FA multimodal neuroimaging data of brain and spinal cord. This is to facilitate determining disease progression, severity, differences among age groups and finally the relationship between imaging biomarkers and clinical/cognitive/blood test measures. There would be two aspects of this project: 1. Development of domain-specific AI models on the macro-micro measures derived from the imaging data 2. Development of Deep learning classification/regression/segmentation models for the identification of novel biomarkers and their respective locations using the image inputs directly without requiring any manual intervention/intermediary steps. This is particularly useful when significant morphological changes in brain/spinal cord make it challenging to derive the automated secondary measures. 3. Development of fusion models by combining the image, patient data and clinical outcome measures. We are seeking a PhD student with a background in Computer/Electrical and Electronics/Biomedical Engineering or in Neuroscience/Data-Science, who would be successful at securing a Monash University PhD stipend scholarship and will work on the various artificial intelligence applications on TRACK-FA data as part of their research thesis. A top-up of $5000 per year for 3.5 years will be awarded to a PhD candidate who secures a Monash University scholarship. Some prior knowledge of image analyses would be advantageous but not mandatory. However, coding literacy in python/bash/R and basic knowledge in AI models are mandatory. Masters by Research Degree holders will get priority.