Soft-Label Anonymous Gastric X-ray Image Distillation

TL;DR

Experimental results show that the proposed method can not only effectively compress the medical dataset but also anonymize medical images to protect the patient’s private information.

Abstract

This paper presents a soft-label anonymous gastric X-ray image distillation method based on a gradient descent approach. The sharing of medical data is demanded to construct high-accuracy computer-aided diagnosis (CAD) systems. However, the large size of the medical dataset and privacy protection are remaining problems in medical data sharing, which hindered the research of CAD systems. The idea of our distillation method is to extract the valid information of the medical dataset and generate a tiny distilled dataset that has a different data distribution. Different from model distillation, our method aims to find the optimal distilled images, distilled labels and the optimized learning rate. Experimental results show that the proposed method can not only effectively compress the medical dataset but also anonymize medical images to protect the patient's private information. The proposed approach can improve the efficiency and security of medical data sharing.

Figures (3)

• Figure 1: Examples of full gastric X-ray images: (a) a sample of non-gastritis image, (b) a sample of gastritis image.
• Figure 2: Examples of patch-based gastric X-ray images: (a) irrelevant patches in $\mathcal{I}$, (b) non-gastritis patches in $\mathcal{N}$, (c) gastritis patches in $\mathcal{P}$.
• Figure 3: Examples of distilled images. I: the distilled image of label $\mathcal{I}$, N: the distilled image of label $\mathcal{N}$, P: the distilled image of label ${\mathcal{P}}$, LR: the optimized learning rate.