Efficient Active Learning for Image Classification and Segmentation using a Sample Selection and Conditional Generative Adversarial Network
This addresses data scarcity in medical imaging for researchers and practitioners, though it appears incremental as it combines existing techniques (active learning, cGANs, Bayesian networks).
The authors tackled the problem of limited medical image data for training deep learning systems by proposing an active learning framework that selects informative samples using a Bayesian neural network and generates additional images with conditional GANs. Their method achieved state-of-the-art performance using only about 35% of the full dataset, saving significant time and effort.
Training robust deep learning (DL) systems for medical image classification or segmentation is challenging due to limited images covering different disease types and severity. We propose an active learning (AL) framework to select most informative samples and add to the training data. We use conditional generative adversarial networks (cGANs) to generate realistic chest xray images with different disease characteristics by conditioning its generation on a real image sample. Informative samples to add to the training set are identified using a Bayesian neural network. Experiments show our proposed AL framework is able to achieve state of the art performance by using about 35% of the full dataset, thus saving significant time and effort over conventional methods.