Annotation-Efficient Learning for Medical Image Segmentation based on Noisy Pseudo Labels and Adversarial Learning
This work addresses the critical problem of expensive manual annotation for medical image segmentation, offering a solution for researchers and clinicians by enabling deep learning models to be trained without direct human annotations, thus reducing costs and accelerating research.
This paper proposes an annotation-efficient learning framework for medical image segmentation that avoids manual annotations by generating pseudo labels using an improved Cycle-Consistent Generative Adversarial Network (GAN). The framework incorporates a VAE-based discriminator and a Discriminator-guided Generator Channel Calibration (DGCC) module to produce high-quality pseudo labels, and a noise-robust iterative learning method with noise-weighted Dice loss to handle their inherent noise. The method achieves segmentation performance comparable to or even exceeding that of learning from human annotations across various medical imaging tasks.
Despite that deep learning has achieved state-of-the-art performance for medical image segmentation, its success relies on a large set of manually annotated images for training that are expensive to acquire. In this paper, we propose an annotation-efficient learning framework for segmentation tasks that avoids annotations of training images, where we use an improved Cycle-Consistent Generative Adversarial Network (GAN) to learn from a set of unpaired medical images and auxiliary masks obtained either from a shape model or public datasets. We first use the GAN to generate pseudo labels for our training images under the implicit high-level shape constraint represented by a Variational Auto-encoder (VAE)-based discriminator with the help of the auxiliary masks, and build a Discriminator-guided Generator Channel Calibration (DGCC) module which employs our discriminator's feedback to calibrate the generator for better pseudo labels. To learn from the pseudo labels that are noisy, we further introduce a noise-robust iterative learning method using noise-weighted Dice loss. We validated our framework with two situations: objects with a simple shape model like optic disc in fundus images and fetal head in ultrasound images, and complex structures like lung in X-Ray images and liver in CT images. Experimental results demonstrated that 1) Our VAE-based discriminator and DGCC module help to obtain high-quality pseudo labels. 2) Our proposed noise-robust learning method can effectively overcome the effect of noisy pseudo labels. 3) The segmentation performance of our method without using annotations of training images is close or even comparable to that of learning from human annotations.