Few-Shot Fingerprinting Subject Re-Identification in 3D-MRI and 2D-X-Ray

TL;DR

This paper presents a method for subject re-identification across different medical imaging datasets using few-shot learning with a ResNet-50, achieving high accuracy in MRI and X-ray data.

Contribution

It introduces a subject fingerprinting approach with a triplet loss trained ResNet-50 for effective re-identification in medical images, addressing data leakage issues.

Findings

High re-identification accuracy on ChestXray-14 and BraTS-2021 datasets.

Effective in both standard and challenging few-shot scenarios.

Demonstrates robustness across different imaging modalities.

Abstract

Combining open-source datasets can introduce data leakage if the same subject appears in multiple sets, leading to inflated model performance. To address this, we explore subject fingerprinting, mapping all images of a subject to a distinct region in latent space, to enable subject re-identification via similarity matching. Using a ResNet-50 trained with triplet margin loss, we evaluate few-shot fingerprinting on 3D MRI and 2D X-ray data in both standard (20-way 1-shot) and challenging (1000-way 1-shot) scenarios. The model achieves high Mean- Recall-@-K scores: 99.10% (20-way 1-shot) and 90.06% (500-way 5-shot) on ChestXray-14; 99.20% (20-way 1-shot) and 98.86% (100-way 3-shot) on BraTS- 2021.