Revolutionising epilepsy care with AI/ML
summary
The Australian Epilepsy Project runs a production pipeline on AWS that cuts pre‑surgical assessment time by 70% reduction in neurologist assessment time by auto‑processing MRI, extracting language activation from task fMRI and flagging lesions. The stack uses S3/EFS/RDS, Amazon Textract for medical histories, PGVector for embeddings, Amazon Bedrock (Claude 3.5, Mistral) and Amazon SageMaker for seizure prediction and auto clinical letter generation. Outputs are clinician‑facing reports and a portal that neurologists validate and use.
AI-written summary · about
chapters (9)
- 0:00Introduction to AI in epilepsy care
- 3:13Australian Epilepsy Project overview
- 7:29Building data foundations for AI
- 12:10Data acquisition and processing pipeline
- 18:35AI use cases and AWS services
- 21:12Language laterality detection with ML
- 26:15Medical history explorer with LLMs
- 28:02Patient summaries via AI
- 29:27Closing and Q&A
related talks
18mUsing AI to automate clinical workflows
59mArtificial intelligence in Healthcare
27mTransforming pharmaceutical R&D through industrialized machine learning
22mLeveraging AI/ML for Transformative Software Solutions in Healthcare
31mBig data for tiny patients: Applying ML to pediatrics
17mAI Innovation in Health
26mAccelerating innovation in the healthcare and life sciences industry
28mAI in healthcare - Deep learning for precision psychiatry