LASIGE researcher Nuno Rodrigues and integrated member Sara Silva published “Effective reduction of unnecessary biopsies through a deep-learning-assisted aggressive prostate cancer detector” in the top 10% journal Scientific Reports. The paper resulted from the collaboration with Champalimaud Foundation (Pprtugal), FORTH (Greece), FPO-IRCCS (Italy), and The ProCAncer-I Consortium.

The work presents a hybrid computer-aided diagnosis system combining radiologists and a deep-learning automatic lesion detection model for prostate cancer. While delineating lesions requires a high degree of experience and expertise from the radiologists, computational models can leverage the information from biparametric MRI to locate the lesions with a high degree of certainty. Focusing only on samples with aggressive disease, the outputs of such models can also be seen as an indication for biopsy, effectively reducing unnecessary biopsy screenings. Through a simulated clinical feasibility scenario, a reduction of approximately 20% of unnecessary biopsies was achieved, with a prospective validation showing that this does not lead to a reduction in the number of detected prostate cancer cases.

The paper is available here.