[PDF]

Design a tiny machine learning model to detect epileptic seizures on wearables


Loic Lorente Lemoine

10/05/2024

Supervised by Nhat (Nick) Pham; Moderated by Irena Spasic

The wearable healthcare market has been experiencing significant growth in recent years, reaching £100 million in 2020 and over £60 billion globally by 2028. Healthcare wearable devices are predicted to be the next generation of personal telemedicine practice. This is especially important for patients with chronic diseases and after surgery, where constant monitoring is essential to prevent fatalities. In this project, we will explore how to implement a tiny and efficient machine-learning model to detect epileptic seizure events that are deployable on wearable devices such as smart watches or smart earphones. The project has three tasks: (1) learn how to design a tiny Machine Learning model (TinyML) with Google CoLab and TensorFlow Lite Microcontroller, (2) design a TinyML prediction model to detect seizure events (seizure dataset will be provided), and (3) deploy the designed model on an Arduino platform (will be provided).


Initial Plan (04/02/2024) [Zip Archive]

Final Report (10/05/2024) [Zip Archive]

Publication Form