Using Machine Learning to Investigate Potential Measures of Disease Progression in Parkinson's Disease Using Open Data

Research areas


An innovative study, based in the United States, is using a smartphone app called MPower to track patients with Parkinson's disease and control participants in tasks which could be informative about the disease, such as finger tapping, voice recording, balance and walking. Crucially, the data for this project are publicly available to researchers who have completed the required accreditation process; my PhD student and I have completed this process and have registered a research project under the theme listed above, but we would like to collaborate with an interested RSCS student and/or machine learning research specialist to gain some traction on this massive dataset. 


Research questions could include: 

  • For the postural measures (standing and walking), can the signal be improved by de-noising algorithms? Does this improve test-retest reliability for standard measures? 
  • Which measures (or combination of measures) best discriminate PD patients and age-matched controls? 
  • Which measures (or combination of measures) best discriminate older from younger participants? 
  • Which measures (or combination of measures) change in a predictable way over time for PD participants? 
  • If we find a useful set of these measures, can we replicate this result in our own sample of PD patients and age-matched controls? (NB there is potential to do this across two sites - both our Canberra Hospital lab and the lab of our collaborators at Stanford University, Professor Helen Bronte-Stewart and her lab). 


A background in machine learning is required, as well as programming skills in MATLAB and/or Python. An understanding of research design and statistics would be a distinct advantage, as would a strong record of teamwork and a sense of humour.  

Background Literature

Bot, B. M., Suver, C., Neto, E. C., Kellen, M., Klein, A., Bare, C., et al. (2016). The mPower study, Parkinson disease mobile data collected using ResearchKit. Scientific Data, 3, 160011.

Trister, A. D., Dorsey, E. R., & Friend, S. H. (2016). Smartphones as new tools in the management and understanding of Parkinson’s disease. Npj Parkinson's Disease, 2, 16006.
Palmerini, L., Rocchi, L., Mellone, S., Valzania, F., & Chiari, L. (2011). Feature selection for accelerometer-based posture analysis in Parkinson's disease. IEEE Transactions on Information Technology in Biomedicine, 5(3), 481–490.


The student will develop an understanding of the applications of machine learning to human physiological and medical data, which is a rapidly growing field with many opportunities for new developments and discoveries. There is also potential for inclusion on published papers and conference travel to present them if the student's contribution is sufficient. Further involvement in our ongoing project on Parkinson's disease in collaboration with Stanford University and TCH is also a possibility.

Updated:  1 June 2019/Responsible Officer:  Dean, CECS/Page Contact:  CECS Marketing