AI pathology
PRET, developed at the Hong Kong University of Science and Technology, uses in-context learning to recognise 18 cancer types and complete new pathology tasks with a few annotated slides.
Trained on 500,000 pathology slides, the AI predicts cancer type, treatment response, and survival with over 95% accuracy within minutes.
The company said the AI integration has improved the biomarker detection of the Guardant cancer protein test by over 20%.
It plans to secure more regulatory approvals for its AI pathology platform.
Also, annalise.ai has made key executive appointments to support its global expansion.