Domain-Robust Mitotic Figure Detection with Style Transfer
This addresses domain shift issues in medical imaging for pathologists, but it is incremental as it builds on existing style transfer methods.
The paper tackles domain generalization in mitotic figure detection by proposing a training scheme that uses style transfer to expand style variance across different scanners, achieving positive performance on a test set with unseen scanners.
We propose a new training scheme for domain generalization in mitotic figure detection. Mitotic figures show different characteristics for each scanner. We consider each scanner as a 'domain' and the image distribution specified for each domain as 'style'. The goal is to train our network to be robust on scanner types by using various 'style' images. To expand the style variance, we transfer a style of the training image into arbitrary styles, by defining a module based on StarGAN. Our model with the proposed training scheme shows positive performance on MIDOG Preliminary Test-Set containing scanners never seen before.