Inflammatory auto-immune diseases are highly prevalent and have limited treatment opportunities. Fluorescence optical imaging (FOI) is a new method that enables radiation-free diagnosis of such diseases in hands with results competitive to MRI. However, the analysis of FOI data is currently performed manually.

Here we aim to develop an unsupervised deep-learning approach using variational auto-encoders, that help to understand the diseases in general. Furthermore we aim to diagnose individual patients and assess their progression during treatment.