New AI tool predicts cancer spread with surprising accuracy
Why it matters: This AI could transform cancer treatment, enabling earlier, more targeted interventions and saving lives.
- Université de Genève researchers discovered that cancer spread isn't random but follows a biological 'program,' identifying gene patterns in colon tumor cells that predict metastasis.
- MangroveGS, an AI model, leverages these genetic signals to predict cancer spread risk with about 80% accuracy, working across multiple cancer types.
- Professor Ariel Ruiz i Altaba explains that cancer should be understood as a distorted developmental process, where reactivated biological programs drive tumor formation and spread.
A groundbreaking AI tool, MangroveGS, developed by the Université de Genève, can predict cancer spread with 80% accuracy by identifying specific gene patterns that signal metastasis. This innovation, based on the understanding that cancer follows a biological 'program' rather than being random, could revolutionize treatment decisions and lead to more personalized patient care.
