Video-based Infantile epileptic Spasms Detection from bOdy Movements using AI
This research aims to develop an AI-based system that can detect Infantile Epileptic Spasms (IES) from smartphone videos. This area is recognized as a top priority because early and accurate diagnosis of epilepsy, especially in infants, is crucial for timely treatment and improved outcomes. There are a few studies which analyse the body movements of the patient from RGB videos for epileptic seizure detection and classification. However, the videos were recorded in a controlled environment (e.g. in hospital) and only a small patient cohort (eg <10) was recruited. In contrast, the approach we propose focuses on analysing videos recorded by parents/carers of infants in uncontrolled environments (e.g. home) which will more accurately reflect the community and clinical resources worldwide.
The research will involve organising Patient and public involvement (PPI) activities to co-create a user-centric AI system for supporting caregivers in recording videos with the required quality. The videos will be collected using the secure NHS-adopted vCreate Health Web-App, which will be assessed and annotated by clinical experts to identify body movement patterns associated with IES. An AI software prototype will be developed and trained using the video data with the clinical annotation to perform the automated IES detection. The data access and analysis will be done within a secured cloud computing platform to ensure data security and privacy. The AI prototype to be developed will serve as the pilot study results to inform our future research and development directions for patient pathway integration.
News and Events
📢 We’re hiring a Postdoctoral Research Associate!
We’re looking for a Postdoctoral Research Associate to join the VISDOM-AI project (Detecting Infantile Epileptic Spasms from smartphone videos), funded by the Epilepsy Research Institute. 🧠 About the …
Read More📢 We’re hiring 2x Research Assistants!
We’re looking for part-time RAs (0.2FTE x 2) to join the VISDOM-AI project (Detecting Infantile Epileptic Spasms from smartphone videos), funded by the Epilepsy Research Institute. 🧠 About the role…
Read MoreThe launch of VISDOM-AI
Happy to share that our team has secured research funding from the Epilepsy Research Institute Endeavour Project Grant to launch a new project – ‘VISDOM-AI’! [official announcement] …
Read MorePresenting at Healthcare Innovation (HI) Scotland Innovation Showcase Day
Sameer, Edmond and Ben gave a talk on the ‘BrAInwaves’ project at the Healthcare Innovation (HI) Scotland Innovation Showcase Day at the Royal Hospital for Children to introduce our video-…
Read MoreOrganized Co-located workshop at BPNA in Oxford
Sameer and Edmond organised a co-located workshop at the British Paediatric Neurology Association (BPNA) Annual Conference in Oxford for clinicians to discuss our EPSRC/MRC IAA project on Cloud-based …
Read MorePPI – Visiting Doune Primary School
On 1st November 2023, Edmond, Sameer, Neil and Isla co-organized a workshop in collaboration with the Scottish Children’s Parliament, Alan Turing Institute, and Scottish AI Alliance on “Exploring Chil…
Read MoreOur Team
Dr. Edmond S. L. Ho
Principal Investigator (PI)
Senior Lecturer in machine learning
university of Glasgow
Prof. Sameer Zuberi
Co-Investigator (Co-I)
Consultant Paediatric Neurologist & Honorary Professor in Paediatric Neurology
Royal Hospital for Children & University of Glasgow
Dr. Neil Patel
CO-Investigator (Co-I)
consultant NeonatologisT & Honorary Associate Professor
Royal Hospital for Children & University of Glasgow
Ben Moore
Project Partner
Founder
vCreate Ltd
Isla Birnie
Project ManageR
Research & Learning Manager
University of Glasgow
Acknowledgement
The project was supported in part by the Epilepsy Research Institute Endeavour Project Grant.
Contact Us
email: visdom-ai@aim-hai.org