Prostate-Specific Membrane Antigen Positron Emission Tomography/Computed Tomography-Derived Radiomic Models in...

Area of Science:

• Radiology
• Oncology
• Medical Imaging Analysis

Background:

• Prostate cancer (PC) management benefits from precise risk stratification.
• Prostate-specific membrane antigen (PSMA) positron emission tomography and computed tomography (PET/CT) offers advanced imaging capabilities.
• Radiomics, extracting quantitative imaging features, is emerging for medical data analysis.

Purpose of the Study:

• To review current literature on PSMA PET/CT-derived radiomics for PC risk stratification.
• To assess the performance of radiomic models in predicting PC diagnosis, pathology, and treatment outcomes.

Main Methods:

• Systematic literature search of publications from 2017 to 2023.
• Analysis of 23 studies focusing on PSMA PET/CT-derived radiomics in prostate cancer.
• Evaluation of reported performance metrics, including receiver operator characteristic curve area under the curve (ROC-AUC).

Main Results:

• PSMA PET/CT radiomics showed strong performance in PC diagnosis (ROC-AUC 0.85-0.925).
• Models accurately predicted biopsy and surgical pathology (ROC-AUC 0.719-0.95).
• Prediction of recurrence, progression, and survival yielded ROC-AUC values from 0.698-0.90.

Conclusions:

• PSMA PET/CT-derived radiomics demonstrates significant potential for PC risk stratification.
• Current applications are promising but require further validation, especially regarding feature reproducibility and biological significance.
• Rigorous validation is essential before widespread clinical adoption.