Can Machine Learning Models Detect and Predict Lymph Node Involvement in Prostate Cancer? A Comprehensive Systematic...

Area of Science:

• Urology
• Radiology
• Artificial Intelligence

Background:

• Prostate cancer (PCa) management faces challenges with lymph node involvement (LNI) detection.
• Artificial Intelligence (AI) is emerging as a tool to improve LNI assessment in PCa.
• Reducing surgical risks and enhancing patient outcomes are key drivers for AI in PCa LNI analysis.

Purpose of the Study:

• To review and analyze existing studies on AI for PCa LNI detection and prediction.
• To evaluate the initial findings and potential of AI models in PCa LNI assessment.
• To identify limitations and future research directions for AI in PCa LNI.

Main Methods:

• Systematic literature search of MEDLINE databases by two independent reviewers.
• Inclusion of 16 studies investigating AI's role in PCa LNI.
• Methodological quality appraisal using the Radiomics Quality Score.

Main Results:

• AI models using Magnetic Resonance Imaging (MRI) demonstrated comparable LNI prediction accuracy to standard nomograms.
• AI models applied to Computed Tomography (CT) and Positron Emission Tomography (PET)-CT showed high diagnostic and prognostic performance.
• AI exhibits potential for accurate detection and prediction of lymph node metastasis in prostate cancer.

Conclusions:

• AI models present a promising approach for lymph node metastasis detection and prediction in prostate cancer.
• Current limitations include retrospective study designs, lack of standardization, manual segmentation, and small sample sizes.
• Further research is essential to optimize AI tools and validate their clinical utility in PCa LNI assessment.