4 часы - перевести

Physics Guided AI Predicts Cervical Cancer Progression

Researchers have built an AI system that predicts how cervical cancer will progress by combining physics with deep learning. The new framework, called Physics-Informed Neural Surrogate Learning, uses equations for tumour immune and viral dynamics to guide predictions and track how treatment affects tumour growth over time. It was tested on patient data and reached a time-dependent accuracy score of 91 percent for forecasting progression across different stages. Unlike black-box models, the system is designed to stay biologically plausible and explainable to doctors. The study was published in Scientific Reports and focuses on personalised treatment planning. Doctors could use such forecasts to choose timing and intensity of therapy for each patient. The team says more validation is needed before use in clinics.

image
advertisement
"We will never ask you to login, or register through a third-party advertisement."