Clinically Significant Prostate Cancer Detection in bpMRI
A. Saha, M. Hosseinzadeh, H. Huisman (2021), "End-to-End Prostate Cancer Detection in bpMRI via 3D CNNs: Effect of Attention Mechanisms, Clinical Priori and Decoupled False Positive Reduction", Medical Image Analysis: 102155.
A. Saha, M. Hosseinzadeh, H. Huisman (2020), "Encoding Clinical Priori in 3D Convolutional Neural Networks for Prostate Cancer Detection in bpMRI", Medical Imaging Meets NeurIPS Workshop – 34th Conference on Neural Information Processing Systems (NeurIPS), Vancouever, Canada.
Description: This algorithm is a deep learning-based computer-aided detection and diagnosis model (Dual-Attention 3D U-Net), which was trained and validated using 1950 prostate bi-parametric (bpMRI) scans paired with PI-RADS v2 annotations. Its intended use is for research purposes only. Its input requires 3 axial/transverse MRI volumes in
- T2-weighted imaging (T2W)
- High b-value diffusion-weighted imaging (HBV)
- Apparent diffusion coefficient map (ADC)
In return, the model generates a single, probabilistic heatmap for the likelihood of clinically significant prostate cancer, as its output prediction. While the input images (T2W, HBV, ADC) can be of different spatial resolutions, the algorithm assumes that they are co-registered or aligned reasonably well, and that the prostate gland is localized within a volume of 336 cm³ from their center coordinate.