Learn the new, keep the old: Extending pretrained models with new anatomy and images

Abstract

Deep learning has been widely accepted as a promising solution for medical image segmentation, given a sufficiently large representative dataset of images with corresponding annotations. With ever increasing amounts of annotated medical datasets, it is infeasible to train a learning method always with all data from scratch. This is also doomed to hit computational limits, e.g., memory or runtime feasible for training. Incremental learning can be a potential solution, where new information (images or anatomy) is introduced iteratively. Nevertheless, for the preservation of the collective information, it is essential to keep some "important" (i.e. representative) images and annotations from the past, while adding new information. In this paper, we introduce a framework for applying incremental learning for segmentation and propose novel methods for selecting representative data therein. We comparatively evaluate our methods in different scenarios using MR images and validate the increased learning capacity with using our methods.

arXiv commentAccepted to MICCAI 2018

References in corpus (4)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Distilling the Knowledge in a Neural Network
• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• Learning by Association - A versatile semi-supervised training method for neural networks

Cited by in corpus (12)

• Uncertainty-aware Contrastive Distillation for Incremental Semantic Segmentation
• Extending Pretrained Segmentation Networks with Additional Anatomical Structures
• Domain-incremental Cardiac Image Segmentation with Style-oriented Replay and Domain-sensitive Feature Whitening
• Active Learning for Segmentation by Optimizing Content Information for Maximal Entropy
• Replay in Deep Learning: Current Approaches and Missing Biological Elements
• Modeling the Background for Incremental and Weakly-Supervised Semantic Segmentation
• Tackling Catastrophic Forgetting and Background Shift in Continual Semantic Segmentation
• Half-Real Half-Fake Distillation for Class-Incremental Semantic Segmentation
• Domain Adaptive Medical Image Segmentation via Adversarial Learning of Disease-Specific Spatial Patterns
• An EM Framework for Online Incremental Learning of Semantic Segmentation
• Mask-guided sample selection for Semi-Supervised Instance Segmentation
• Continual Active Learning for Efficient Adaptation of Machine Learning Models to Changing Image Acquisition