MRI-based radiomics for prediction of biochemical recurrence in prostate cancer: a systematic review and meta-analysis
- 1Research Center of Thoracic Oncology (RCTO), National Research Institute of Tuberculosis and Lung Diseases (NRITLD), Shahid Beheshti University of Medical Sciences, Tehran, Iran. mohsenwork98@gmail.com.
- 2Western Michigan University Homer Stryker M.D. School of Medicine, Kalamazoo, MI, USA.
- 3Isfahan Neurosciences Research Center, Isfahan University of Medical Sciences, Isfahan, Iran.
- 4Department of Radiology, Hospital of University of Pennsylvania, University of Pennsylvania, Philadelphia, PA, 19104, USA.
- 5Research Center of Thoracic Oncology (RCTO), National Research Institute of Tuberculosis and Lung Diseases (NRITLD), Shahid Beheshti University of Medical Sciences, Tehran, Iran. sh_seifi@yahoo.com.
- 0Research Center of Thoracic Oncology (RCTO), National Research Institute of Tuberculosis and Lung Diseases (NRITLD), Shahid Beheshti University of Medical Sciences, Tehran, Iran. mohsenwork98@gmail.com.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.Machine learning models using MRI radiomics show promise for predicting prostate cancer recurrence. Combining radiomics with clinical data offers the best prediction accuracy, though standardization is needed for clinical use.
Area Of Science
- Radiology
- Oncology
- Artificial Intelligence
Background
- Biochemical recurrence (BCR) after prostate cancer (PCa) treatment predicts metastasis and mortality.
- Early BCR prediction is crucial for guiding treatment and optimizing patient management.
- Magnetic Resonance Imaging (MRI) is vital for PCa diagnosis and surveillance.
Purpose Of The Study
- To evaluate the accuracy and quality of MRI radiomics-based machine learning (ML) models for predicting post-treatment BCR in PCa.
- To systematically review and meta-analyze existing studies on this topic.
Main Methods
- A systematic literature search was performed across five databases up to December 23, 2024.
- Studies were assessed for quality using QUADAS-2 and METRICS tools.
- A meta-analysis of radiomics, clinical, and clinical-radiomics models was conducted to pool performance metrics (sensitivity, specificity, AUC).
Main Results
- Twenty-four studies were reviewed; 14 were included in the meta-analysis.
- Radiomics-based ML models achieved a pooled AUC of 0.75, sensitivity of 72%, and specificity of 78%.
- Clinical-radiomics models demonstrated superior performance with a pooled AUC of 0.88, sensitivity of 85%, and specificity of 79%.
Conclusions
- MRI radiomics shows potential for predicting PCa BCR, particularly when combined with clinical data.
- Widespread clinical adoption requires further standardization and methodological improvements for generalizability.
- Future research should focus on multi-center designs and external validation to enhance clinical applicability.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.