Radiomics and Clinicopathological Characteristics for Predicting Lymph Node Metastasis in Testicular Cancer.

Area of Science:

• Oncology
• Radiology
• Medical Imaging

Background:

• Accurate prediction of lymph node metastasis (LNM) in testicular cancer is vital for treatment planning and prognosis.
• Current methods may not fully capture all predictive information for LNM.

Purpose of the Study:

• To develop and validate clinical radiomics models for preoperative prediction of LNM in testicular cancer.
• To identify the most effective machine learning approach for this prediction task.

Main Methods:

• Utilized data from 91 early-stage testicular cancer patients.
• Integrated clinical risk factors (age, tumor markers, histotype, BMI) with retroperitoneal lymph node radiomics features.
• Developed predictive models using Random Forest (RF), Light Gradient Boosting Machine (LGBM), Support Vector Machine Classifier (SVC), and K-Nearest Neighbours (KNN).
• Assessed model performance using Area Under the Receiver Operating Characteristic Curve (AUC) and clinical utility via Decision Curve Analysis (DCA).

Main Results:

• The combined Random Forest model achieved the highest AUC of 0.95 (±0.03 SD).
• Decision Curve Analysis confirmed the clinical usefulness of the candidate model for preoperative prediction.
• The study identified reliable machine learning techniques for LNM prediction.

Conclusions:

• Machine learning integrating clinical factors and radiomics offers a powerful tool for predicting LNM in testicular cancer.
• This approach enhances precision oncology by improving preoperative risk stratification.
• The findings support the expanded application of radiomics in cancer treatment decision-making.