Predictive analytics

A neurologist's startup has developed a foundation model, based on 60,000 hours of brain wave records, which reportedly achieved 95% accuracy in identifying epilepsy biomarkers.

Shlomi Madar, CEO of SpotitEarly, discusses the company's noninvasive test using a breath sample, which is analyzed by specially trained dogs capable of identifying unique cancer-related scent signatures combined with AI technology.

Seoul National University Hospital researchers have developed an AI model that predicts the response to an anticonvulsant drug.

Sam Davis Jr., in the second part of our interview, says meeting regularly with surgical staff and building trust in its predictive analytics helped Rush University Medical Center enhance OR efficiency and forecast surgical demand with 90% accuracy.

Sam Davis Jr. of Rush University Medical Center says adopting analytics enabled the system to align staffing, equipment and surgeon block time with real demand, delivering 12x ROI and improving patient outcomes.

The model can also make zero-shot predictions.

The multi-task model, trained on 80,000 patient records, predicts three key postoperative risks and validates consistently across several hospitals.

The model also calculates biological ageing and survival patterns to support personalized health risk management.

The acquisition expands Datavant's capabilities to help insurance and legal professionals with AI-enabled tools for automated medical record review and predictive analytics.

Elsa, the FDA's new AI tool, promises to speed up safety reviews. ConcertAI's Jeff Elton advises developers to anticipate its insights and optimize interactions by preparing structured, contextualized data.