Related Experiment Video
Updated: Jun 19, 2025
07:15
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
6.8K
Development and Validation of a Biparametric MRI Deep Learning Radiomics Model with Clinical Characteristics for
Yue-Yue Zhang1, Hui-Min Mao2, Chao-Gang Wei3
1Department of Radiology, Children's Hospital of Soochow University, Suzhou 215025, China; Department of Radiology, Second Hospital of Soochow University, Suzhou 215004, China.
Academic Radiology
|July 23, 2024
Summary
A new deep learning model integrating MRI radiomics and clinical data accurately predicts perineural invasion (PNI) in prostate cancer (PCa). This tool aids in developing personalized treatment strategies for PCa patients.
Area of Science:
- Oncology
- Radiology
- Artificial Intelligence
Background:
- Perineural invasion (PNI) is a critical prognostic factor in prostate cancer (PCa).
- Accurate non-invasive prediction of PNI is essential for treatment planning.
Purpose of the Study:
- To develop and validate a predictive model for non-invasive PNI detection in PCa.
- Integrate biparametric MRI-based deep learning radiomics and clinical data.
Main Methods:
- Prospective study of 557 PCa patients undergoing MRI and radical prostatectomy.
- Development of clinical, radiomics, deep learning, and integrated (DLRC) models.
- Performance evaluation using ROC and PR curves, calibration, and decision curves.
Main Results:
- The integrated DLRC model achieved high performance in both training and validation cohorts.
- ROC-AUCs of 0.914 (training) and 0.848 (validation).
- PR-AUCs of 0.948 (training) and 0.926 (validation).
Conclusions:
- The DLRC model is a robust tool for predicting PNI in PCa patients.
- This model can assist in formulating effective treatment strategies.
- Non-invasive prediction of PNI improves patient management.
