Design and development of machine learning and signal processing algorithms for a wearable device (Watch HOP) for critical care, surgery, and chronic disease monitoring

Diabetes is a chronic illness characterized by elevated levels of blood glucose, accompanied by disturbed metabolism of fats and proteins. In the present project, a non-invasive wrist-worn device (Watch HOP) will be designed and fabricated to obtain physiological data (digital biomarkers) from participants to enable the creation of an algorithm that will ultimately allow to monitor and predict HbA1c levels and assess the efficacy of the given treatment.
In many applications of the smartwatch, it is desirable to record data for long periods of time. As a result, the recoded data will have a large dimension. Therefore, compression of the signal is an important part of design process. Compression algorithms, represent the original signal with fewer information bits. It is important to preserve all the information in the original signal in the compression method. The candidate will work with the hardware and firmware team to design and implement a lossless compression algorithm to compress the recorded signals.

Faculty Supervisor:

Soosan Beheshti

Student:

Partner:

HOP Technologies

Discipline:

Engineering

Sector:

Manufacturing; Professional, scientific and technical services

University:

Toronto Metropolitan University

Program: