
This study aimed to assess the feasibility of using an AI algorithm alongside clinical data for diagnosing clinically significant prostate cancer (csPCa) from multiparametric MRI (mpMRI) results. A retrospective cohort of 505 patients was analyzed, considering age, PSA levels, and mpMRI results. An AI algorithm processed mpMRI images, generating csPCa predictions. Logistic regression models were developed, comparing a conventional model (using age, PSA, and mpMRI scores) and an AI model (with AI predictions). The AI model showed higher prediction efficacy (AUC 0.849) than the conventional model (AUC 0.782), indicating that incorporating AI-generated data alongside clinical information could enhance csPCa diagnosis.
Like
Save
Share