Statistical machine learning for remote characterization of neurological disorders using smartphones

Add to your list(s) Download to your calendar using vCal

• Dr Max Little (School of Engineering & Applied Science, Aston University)
• Monday 16 April 2018, 11:00-12:00
• Computer Science, The Sloman Lounge (UG).

If you have a question about this talk, please contact Hector Basevi.

Host: Prof Peter Tino

Website: www.maxlittle.net

Abstract: These days smartphones and other wearable devices come packed with movement sensors, microphones and other digital sensors. For diseases such as Parkinson’s and Huntington’s which are movement disorders, we can attempt to use these sensors to do remote symptom monitoring. In collaboration with several groups including Johns Hopkins (Baltimore, US), Oxford and Rochester (NY, US), we have pioneered the development of novel statistical machine learning algorithms for processing sensor data captured for this purpose. I will describe our efforts to provide interpretable clinical value from this data using specialized algorithms such as regularized piecewise linear trend filtering, infinite HMMs with AR observations and weakly-supervised maximum margin regression to process the digital sensor data. I will describe our current efforts in combining these algorithms to provide useful summarization of this data for clinical trials, symptom severity monitoring for disease management, and detecting the earliest symptoms to provide objective data towards early differential diagnosis.

This talk is part of the Artificial Intelligence and Natural Computation seminars series.

This talk is included in these lists:

• Artificial Intelligence and Natural Computation seminars
• Computer Science Departmental Series
• Computer Science Distinguished Seminars
• Computer Science, The Sloman Lounge (UG)

Note that ex-directory lists are not shown.