Supervision by Denoising for Medical Image Segmentation
arXiv:2202.02952 · doi:10.1109/TPAMI.2023.3299789
Abstract
Learning-based image reconstruction models, such as those based on the U-Net, require a large set of labeled images if good generalization is to be guaranteed. In some imaging domains, however, labeled data with pixel- or voxel-level label accuracy are scarce due to the cost of acquiring them. This problem is exacerbated further in domains like medical imaging, where there is no single ground truth label, resulting in large amounts of repeat variability in the labels. Therefore, training reconstruction networks to generalize better by learning from both labeled and unlabeled examples (called semi-supervised learning) is problem of practical and theoretical interest. However, traditional semi-supervised learning methods for image reconstruction often necessitate handcrafting a differentiable regularizer specific to some given imaging problem, which can be extremely time-consuming. In this work, we propose "supervision by denoising" (SUD), a framework that enables us to supervise reconstruction models using their own denoised output as soft labels. SUD unifies stochastic averaging and spatial denoising techniques under a spatio-temporal denoising framework and alternates denoising and model weight update steps in an optimization framework for semi-supervision. As example applications, we apply SUD to two problems arising from biomedical imaging -- anatomical brain reconstruction (3D) and cortical parcellation (2D) -- to demonstrate a significant improvement in the image reconstructions over supervised-only and stochastic averaging baselines.
To appear in the IEEE Transactions on Pattern Analysis and Machine Intelligence
References in corpus (16)
- Automated Design of Deep Learning Methods for Biomedical Image Segmentation
- Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
- mixup: Beyond Empirical Risk Minimization
- Temporal Ensembling for Semi-Supervised Learning
- Deep Learning for Segmentation using an Open Large-Scale Dataset in 2D Echocardiography
- Anatomically Constrained Neural Networks (ACNN): Application to Cardiac Image Enhancement and Segmentation
- SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining
- Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning
- Deep Learning-enabled Virtual Histological Staining of Biological Samples
- Plug-and-Play Priors for Bright Field Electron Tomography and Sparse Interpolation
- A Topological Loss Function for Deep-Learning based Image Segmentation using Persistent Homology
- Learning Proximal Operators: Using Denoising Networks for Regularizing Inverse Imaging Problems
- SynthMorph: learning contrast-invariant registration without acquired images
- Semi-Supervised Medical Image Segmentation via Learning Consistency under Transformations
- FlatNet: Towards Photorealistic Scene Reconstruction from Lensless Measurements
- Deep Retinal Image Segmentation with Regularization Under Geometric Priors