Related Experiment Video
Updated: Jan 15, 2026
06:08
A Cognitive Fusion-guided Prostate Biopsy Using Multiparametric Magnetic Resonance Imaging and Transrectal Ultrasound
Published on: March 21, 2025
1.1K
A 3D Edge-Attention Denoising Diffusion Network for Prostate Segmentation in Puncture Biopsy
IEEE Journal of Biomedical and Health Informatics
|October 6, 2025
Summary
This study introduces a novel 3D network for accurate prostate segmentation in transrectal ultrasound (TRUS) images, improving biopsy precision. The method enhances edge detection and reduces noise for better diagnostic outcomes.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Biomedical Engineering
Background:
- Prostate cancer diagnosis relies on transrectal ultrasound (TRUS) guided biopsy.
- Accurate prostate segmentation in TRUS images is vital for precise biopsies.
- Automated segmentation is challenging due to image noise and artifacts.
Purpose of the Study:
- To develop a highly accurate and generalizable 3D network for prostate segmentation in TRUS images.
- To overcome limitations of current automated segmentation methods in challenging ultrasound data.
Main Methods:
- Proposed a 3D edge-attention denoising diffusion network.
- Incorporated an edge attention denoising U-Net (EAD U-Net) for edge feature extraction.
- Integrated a Kalman fusion module for uncertainty reduction and optimal segmentation estimation.
Main Results:
- Achieved high segmentation accuracy on 1834 3D ultrasound images from two datasets.
- Demonstrated superior performance compared to existing state-of-the-art methods.
- Obtained an average Dice similarity coefficient of 92.92% and 94.0%, and 95th percentile Hausdorff distance of 1.07 mm and 0.77 mm.
Conclusions:
- The proposed network effectively segments prostates in TRUS images, outperforming current techniques.
- The method shows potential for facilitating accurate MRI-TRUS fusion guided prostate biopsies.
- The approach addresses key challenges in automated ultrasound image segmentation.
