Multimodal deep learning predicts positive surgical margins on preoperative prostate MRI

A multimodal deep learning model combining MRI and clinical features predicted positive surgical margins after radical prostatectomy with an AUC of 0.843 (95% CI 0.770-0.911) in an external validation cohort of 155 patients from six centers, outperforming imaging-only and clinic…

• Multicenter retrospective study: 1022 patients from two institutions for training, 155 from six independent centers for external validation.
• Multimodal fusion model (MDESM) AUC 0.843 (95% CI 0.770-0.911) significantly higher than imaging-only DESM (0.711) and clinical-only model (0.676) in external validation.
• Decision curve analysis showed MDESM provided the highest net benefit across threshold probabilities.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.