Hybrid Multi-Dimensional MRI Prostate Cancer Detection via Hadamard Network-Based Bias Correction and Residual Networks
Researchers propose HBR-Net-18, a two-stage neural network combining Hadamard U-Net bias correction with ResNet-18 classification to automate prostate cancer detection from multi-parametric MRI scans. The framework addresses intensity inhomogeneities in tissue composition maps to improve diagnostic accuracy.42
