What happened
Two intensive care physicians at St. Michael's Hospital in Toronto, Dr. Bijan Teja and Dr. Michael Sklar, are developing AI Compass, a tool intended to detect early signs that an ICU patient is about to deteriorate or die, BetaKit reported. The system is to be trained on records from tens of thousands of ICU patients to find patterns clinicians may not recognize, with the goal of giving care teams more time to act.
The project received $200,000 through the Mitchell Award for Health System Innovation in the St. Michael's Hospital Foundation's Angel's Den competition, which bills itself as Canada's largest medical research competition and had $700,000 available in this round. Teja said the money lets the team train the model and move toward a proof-of-concept stage, but that additional funding will be required.
AI Compass is not yet in use in any ICU. Its 30-person multidisciplinary team inside the hospital, which includes Dr. Amol Verma, plans to check retrospectively whether the tool would have identified at-risk St. Michael's patients sooner. Teja and Sklar have worked on the project since roughly 2024, and Teja said his long-term aim is to see it running in hundreds of ICUs in Canada and elsewhere.
Context
St. Michael's already uses ChartWatch, a tool whose development Verma led, which monitors general ward patients and warns when they may need ICU care. A hospital study tied it to a 26-percent decline in unexpected deaths on the general medicine ward. Wards usually record vital signs every four hours, while ICUs capture data every minute, including ventilator and life-support settings, giving AI Compass richer inputs.
ブリーフィング
1週間のテックニュースを1通のメールで。
読む価値のある記事を毎週お届けします。配信停止はいつでも可能です。
