Related Experiment Video
Updated: Jun 26, 2025
06:08
A Cognitive Fusion-guided Prostate Biopsy Using Multiparametric Magnetic Resonance Imaging and Transrectal Ultrasound
Published on: March 21, 2025
169
Development and Validation of a Multimodality Model Based on Whole-Slide Imaging and Biparametric MRI for Predicting
Chenhan Hu1, Xiaomeng Qiao1, Renpeng Huang1
1From the Departments of Radiology (Chenhan Hu, X.Q., Chunhong Hu, J.B., X.W.) and Pathology (R.H.), the First Affiliated Hospital of Soochow University, 188 Shizi Road, Suzhou 215006, China.
Radiology. Imaging Cancer
|May 17, 2024
Summary
A new machine learning model combining MRI, WSI, and clinical data accurately predicts prostate cancer recurrence after surgery. This multimodality approach offers improved prediction for personalized treatment planning.
Area of Science:
- Oncology
- Radiology
- Pathology
Background:
- Prostate cancer (PCa) biochemical recurrence (BCR) after radical prostatectomy (RP) impacts treatment decisions.
- Accurate prediction of BCR is crucial for guiding postoperative management and improving patient outcomes.
Purpose of the Study:
- To develop and validate a machine learning multimodality model for predicting PCa BCR.
- The model integrates preoperative MRI, surgical whole-slide imaging (WSI), and clinical variables.
Main Methods:
- A retrospective study included 363 male patients with PCa undergoing RP.
- A least absolute shrinkage and selection operator Cox algorithm selected clinical variables.
- Radiomics and pathomics signatures were derived from MRI and WSI data, respectively.
- A multimodality model combined radiomics, pathomics, and clinical signatures.
Main Results:
- Radiomics and pathomics signatures showed good predictive performance for BCR (C-index: 0.742 and 0.730).
- The multimodality model achieved the highest predictive performance (C-index: 0.860) in the testing cohort.
- This performance was significantly superior to single-modality models (P <= .01).
Conclusions:
- The developed multimodality model accurately predicts BCR after RP in PCa patients.
- This model serves as a promising tool for assisting in postoperative individualized treatment strategies.
