RadHop-Net: A Lightweight Radiomics-to-Error Regression for False Positive Reduction In MRI Prostate Cancer Detection

2025-01-03

AI Summary

To address the high false-positive rate in bpMRI-based prostate cancer detection—leading to unnecessary biopsies and increased patient burden—this paper proposes a two-stage lightweight CNN framework. In Stage I, candidate lesions are identified via data-driven radiomics. In Stage II, RadHop-Net performs lesion-level classification using adaptive receptive field expansion and error-regression–based bias correction. We introduce the novel “radiomics-to-error-regression” paradigm, moving beyond conventional voxel-to-label mapping. A weighted regression loss function is designed to explicitly balance false-positive and true-positive penalties. The model achieves state-of-the-art performance with significantly fewer parameters: mean average precision (AP) improves from 0.407 to 0.468 on the PI-CAI dataset, while model size is drastically reduced—enabling high accuracy alongside clinical deployability.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Multi-instance/Multi-view Learning
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Clinically significant prostate cancer (csPCa) is a leading cause of cancer death in men, yet it has a high survival rate if diagnosed early. Bi-parametric MRI (bpMRI) reading has become a prominent screening test for csPCa. However, this process has a high false positive (FP) rate, incurring higher diagnostic costs and patient discomfort. This paper introduces RadHop-Net, a novel and lightweight CNN for FP reduction. The pipeline consists of two stages: Stage 1 employs data driven radiomics to extract candidate ROIs. In contrast, Stage 2 expands the receptive field about each ROI using RadHop-Net to compensate for the predicted error from Stage 1. Moreover, a novel loss function for regression problems is introduced to balance the influence between FPs and true positives (TPs). RadHop-Net is trained in a radiomics-to-error manner, thus decoupling from the common voxel-to-label approach. The proposed Stage 2 improves the average precision (AP) in lesion detection from 0.407 to 0.468 in the publicly available pi-cai dataset, also maintaining a significantly smaller model size than the state-of-the-art.

Problem

• Prostate Cancer
• MRI
• False Positives Reduction

Innovation

• RadHop-Net
• bpMRI
• Prostate Cancer Detection