Applying Machine Learning to Healthcare

In the last blog we announced that Mohamed Bennasar will be joining us- this is some more background about his research experience and how he hopes to apply it to STRETCH.

I have been in the area of machine learning, and signal and image processing since 2010, my main research interests are: pattern recognition, and machine learning. I have employed these techniques for health care application; to help clinicians in diagnosing, and monitoring the progression of diseases.

Machine Learning (ML) based systems have become a part of modern life, they are being used for example in mail delivery, reading car plates, search engines, and translators. As a part of my PhD, I developed a computerised clinical decision support system for early detection of dementia based on the clock drawing test. This system uses Artificial Intelligence (AI) techniques to analyse the clock drawings and classify them into: healthy, dementia, or early stage of dementia.  The system achieved high accuracy in diagnosing positive dementia cases.

I worked as a Research Associate in Cardiff University when I developed a system for assessing the progression of Huntington’s Disease. The system uses sensors to monitor the movements of the patients during a simple transfer task called money box test. Computer algorithms are deployed to analyse the sensors data and generate a score of the degree of the movement impairment.

Working in the Open University as a part of the STRETCH project will give me very good opportunity to combine techniques of ML and the sensor technology to provide tools that can facilitate healthcare support. ML techniques have the capability to discover the underlying relations between data, and study the significance of each part of the data. ML techniques also can be employed to build decision support tools that can assist the people in the circle of support in making decisions.  

Leave a Reply

Your email address will not be published. Required fields are marked *