Adjusting for Confounding in Unsupervised Latent Representations of Images
This addresses the issue of unwanted variability in biological imaging for researchers, but it is incremental as it builds on existing adversarial methods.
The paper tackles the problem of confounding factors in biological imaging data by proposing an adversarial training strategy to learn unsupervised representations invariant to nuisances, and validates it using deep convolutional autoencoders on microscopy imaging.
Biological imaging data are often partially confounded or contain unwanted variability. Examples of such phenomena include variable lighting across microscopy image captures, stain intensity variation in histological slides, and batch effects for high throughput drug screening assays. Therefore, to develop "fair" models which generalise well to unseen examples, it is crucial to learn data representations that are insensitive to nuisance factors of variation. In this paper, we present a strategy based on adversarial training, capable of learning unsupervised representations invariant to confounders. As an empirical validation of our method, we use deep convolutional autoencoders to learn unbiased cellular representations from microscopy imaging.