Goldilocks-curriculum Domain Randomization and Fractal Perlin Noise with Application to Sim2Real Pneumonia Lesion Detection

TL;DR

This paper introduces a new sim2real transfer method using Goldilocks-curriculum domain randomization and fractal Perlin noise, validated on a benchmark dataset for pneumonia lesion detection in chest X-rays.

Contribution

It proposes a novel domain randomization technique tailored for medical imaging, specifically addressing data scarcity in pneumonia detection.

Findings

GDR improves sim2real transfer performance

Constructed a challenging pneumonia X-ray benchmark dataset

Validated the effectiveness of GDR in medical image analysis

Abstract

A computer-aided detection (CAD) system based on machine learning is expected to assist radiologists in making a diagnosis. It is desirable to build CAD systems for the various types of diseases accumulating daily in a hospital. An obstacle in developing a CAD system for a disease is that the number of medical images is typically too small to improve the performance of the machine learning model. In this paper, we aim to explore ways to address this problem through a sim2real transfer approach in medical image fields. To build a platform to evaluate the performance of sim2real transfer methods in the field of medical imaging, we construct a benchmark dataset that consists of $101$ chest X-images with difficult-to-identify pneumonia lesions judged by an experienced radiologist and a simulator based on fractal Perlin noise and the X-ray principle for generating pseudo pneumonia lesions. We then develop a novel domain randomization method, called Goldilocks-curriculum domain randomization (GDR) and evaluate our method in this platform.