Pelvic lymph node invasion in patients with prostate cancer is associated with different treatment selection and planning while there is no clear consensus on nomograms that can be clinically available for prediction of lymph node invasion. Our predictive model, based on preoperative clinical characteristics, MR image features and biopsy findings of 248 consecutive patients, was trained with a support vector machine and compared to a logistic regression analysis, allowing for improved differentiation in assessing the risk of lymph node invasion. Use of this machine-learning-based predictive tool potentially connect to better selection of optimal type of treatment and long-term excellent prognosis.
This abstract and the presentation materials are available to members only; a login is required.