Related Experiment Video
Updated: Mar 18, 2026
06:08
A Cognitive Fusion-guided Prostate Biopsy Using Multiparametric Magnetic Resonance Imaging and Transrectal Ultrasound
Published on: March 21, 2025
1.9K
PRAD-Hybrid CNN (PRADHC): A Deep Learning Model for Assisted Diagnosis of Prostate Cancer on MRI
Jingpeng Liu1, Lingxuan Hou2, Yang Xu3
1Department of Urology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China.
Current Medical Imaging
|March 16, 2026
Summary
A new Prostate Adenocarcinoma Hybrid Convolutional Neural Network (PRADHC) model significantly improves prostate cancer diagnosis using MRI scans. This AI tool enhances accuracy, aiding clinicians in detecting early-stage cancers more effectively.
Area of Science:
- Artificial Intelligence in Medical Imaging
- Deep Learning for Cancer Detection
- Radiology and Urology Innovations
Background:
- Prostate cancer is a common male malignancy.
- Current prostate MRI diagnosis is subjective and may miss early cancers.
- There is a need for more efficient diagnostic techniques.
Purpose of the Study:
- To introduce a novel deep learning model for automated prostate cancer diagnosis.
- To enhance diagnostic accuracy, particularly for early-stage prostate cancer.
- To provide clinicians with an effective tool for assisted diagnosis.
Main Methods:
- Developed the Prostate Adenocarcinoma Hybrid Convolutional Neural Network (PRADHC) model, combining EfficientNet and Residual Blocks.
- Validated the PRADHC model on 1,528 MRI images from 64 patients.
- Enhanced EfficientNet with more CNN layers and integrated Residual Networks to improve accuracy and mitigate gradient vanishing.
Main Results:
- The PRADHC model achieved 99.34% accuracy and 99.32% AUC.
- Demonstrated a 4% improvement in accuracy compared to conventional EfficientNet.
- Outperformed a baseline elementary CNN model (95.72% accuracy, 96.74% AUC).
Conclusions:
- The PRADHC model offers a novel deep learning approach for automated prostate cancer diagnosis.
- This AI tool can augment radiologists' and urologists' diagnostic capabilities, aiding early detection and treatment planning.
- While acknowledging limitations like false positives, the system serves as a valuable supplementary tool for diagnostic support.
