Innovation and data analytics: Solving real-world challenges
summary
UBC CIC teams deliver working prototypes: an ML model on heart‑failure patient records reaches about 70% accuracy after synthesizing and anonymizing messy clinical data; an open‑science pipeline (Serratus) scans public genomes and finds tenfold more RNA viruses in 11 days; a COVID CT ResNet model trains on ~3,000 de‑identified studies and enters clinical trials; an MRI requisition tool uses Amazon Comprehend Medical plus a rule‑editable UI to standardize priorities; Sea Around Us moves fisheries data to S3/Athena/Jupyter and cuts staff query work by ~80%, and TapGives runs USSD water subscriptions with usage analytics.
AI-written summary · about
chapters (12)
- 0:00Introducing the UBC Cloud Innovation Center
- 2:27How the CIC was created
- 5:38CIC process and open source approach
- 9:11Heart failure risk prediction project
- 14:17Sea Around Us data optimization
- 19:39MRI booking prioritization system
- 27:14AI for COVID-19 CT diagnostics
- 33:28Serratus open science project
- 38:48TapGives clean water subscription service
- 43:16Student training and project outcomes
- 46:26Supporting projects after deployment
- 49:16Q&A session
related talks
07mCollaboration in AI benefits humanity
22mAI-powered digital transformation in healthcare and life sciences
03mAI solutions for healthcare
45mDigital Disruption in health & medical research
22mReimaging Healthcare
33mData Science in Healthcare: Beyond the Hype
45mCreando soluciones inteligentes en el sector salud
09mHow AI is changing the healthcare system