Old prescriptions, new predictions.
This is Sandra Tsing Loh with the Loh Down on Science.
Welcome to the Artificial Intelligence Era! AI can now potentially predict melanoma — a skin cancer — before it strikes. How? With help from medical records.
Enter Martin Gillstedt at the University of Gothenburg in Sweden. The team first predicted melanoma using demographic information. Things like age, gender, and region of birth – from over six MILLION adults! Then they added medical history into the AI model.
Results? Adding past diagnoses and prescriptions made the predictions significantly more accurate. The computer spotted high-risk patients it would have missed. Sharper prediction, earlier detection, lives saved!
AI turned your pill bottle into a fortune teller — no more fiction!
Reference: Gillstedt, M., Stempfle, L., Paoli, J., Johansson, F. D., & Polesie, S. (2026). Predicting melanoma impact on the Swedish healthcare system from the adult population using machine learning on registry data. *Acta Dermato-Venereologica, 106*, adv 44610. https://doi.org/10.2340/actadv.v106.44610