Goldilocks-curriculum Domain Randomization and Fractal Perlin Noise with Application to Sim2Real Pneumonia Lesion Detection
This addresses the challenge of limited data for medical imaging CAD systems, though it is incremental as it builds on existing sim2real and domain randomization techniques.
The paper tackled the problem of insufficient medical images for training computer-aided detection systems by developing a sim2real transfer approach, resulting in a benchmark dataset of 101 chest X-rays and a novel domain randomization method that improved performance in pneumonia lesion detection.
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.