Patient foundation model for risk stratification in low-risk overweight patients
This AI could revolutionize how doctors identify high-risk patients and allocate expensive treatments.
Researchers have developed PatientTPP, a new AI foundation model trained on over 500,000 real-world patient records to predict health risks in overweight individuals. The model analyzes sequences of diagnoses, lab results, and medications to create patient representations. It significantly outperformed traditional body mass index (BMI) in identifying patients with higher future cardiovascular-related healthcare costs, offering a more precise tool for preventive care and targeting costly therapies like GLP-1 agonists.
Why It Matters
This enables more accurate, cost-effective preventive healthcare by identifying high-risk patients earlier than current methods.