Multi-task pre-training of deep neural networks for digital pathology
This work addresses the problem of limited data for classification tasks in digital pathology, offering an incremental improvement over existing pre-training methods.
The paper tackles the lack of a large-scale dataset like ImageNet in digital pathology by assembling 22 classification tasks with nearly 900k images and using multi-task pre-training. The results show that their models, when used as feature extractors, either significantly outperform or match ImageNet pre-trained models, with fine-tuning further improving performance to comparable levels.
In this work, we investigate multi-task learning as a way of pre-training models for classification tasks in digital pathology. It is motivated by the fact that many small and medium-size datasets have been released by the community over the years whereas there is no large scale dataset similar to ImageNet in the domain. We first assemble and transform many digital pathology datasets into a pool of 22 classification tasks and almost 900k images. Then, we propose a simple architecture and training scheme for creating a transferable model and a robust evaluation and selection protocol in order to evaluate our method. Depending on the target task, we show that our models used as feature extractors either improve significantly over ImageNet pre-trained models or provide comparable performance. Fine-tuning improves performance over feature extraction and is able to recover the lack of specificity of ImageNet features, as both pre-training sources yield comparable performance.