Decoupled conditional contrastive learning with variable metadata for prostate lesion detection

TL;DR

This paper introduces a novel contrastive learning method that utilizes variable confidence metadata from radiology reports to improve prostate lesion detection in MRI, achieving a 3% AUC increase.

Contribution

It proposes a decoupled conditional contrastive loss that incorporates metadata confidence, addressing variability in radiology report annotations for better lesion detection.

Findings

Achieved 3% AUC improvement on PI-CAI dataset.

Effectively leverages weak and variable metadata in contrastive learning.

Demonstrates robustness to inter-reports variability.

Abstract

Early diagnosis of prostate cancer is crucial for efficient treatment. Multi-parametric Magnetic Resonance Images (mp-MRI) are widely used for lesion detection. The Prostate Imaging Reporting and Data System (PI-RADS) has standardized interpretation of prostate MRI by defining a score for lesion malignancy. PI-RADS data is readily available from radiology reports but is subject to high inter-reports variability. We propose a new contrastive loss function that leverages weak metadata with multiple annotators per sample and takes advantage of inter-reports variability by defining metadata confidence. By combining metadata of varying confidence with unannotated data into a single conditional contrastive loss function, we report a 3% AUC increase on lesion detection on the public PI-CAI challenge dataset. Code is available at: https://github.com/camilleruppli/decoupled_ccl