Cost-effective survival prediction for patients with advanced prostate cancer using clinical trial and real-world hospital registry datasets

Mika Murtojärvi, Anni S. Halkola, Antti Airola, Teemu D. Laajala, Tuomas Mirtti, Tero Aittokallio, Tapio Pahikkala

Abstract: Highlights•Two variable selection methods are implemented: LASSO and greedy selection.•The methods are tested in four cohorts of prostate cancer patients.•The cost of survival prediction for patients with prostate cancer can be significantly reduced.•LASSO tends to give better peak accuracy but with low budgets the greedy method is better.