Reda Kamraoui receives the MICCAI International Challenge Award for his work on the detection of new multiple sclerosis lesions.
The detection of new multiple sclerosis lesions is an important marker of disease progression. The applicability of deep learning-based methods could automate this task. However, no previous work has addressed the problem of sparse longitudinal data and generalization of methods. In this work, we first propose a pipeline to generate two realistic time points from a single MRI sequence. Second, we use image quality data augmentation to improve the generalization capability of our segmentation model. Overall, our contributions lead to the best performance in the MICCAI MSSEG2 Challenge of segmentation of new MS lesions (see Commowick et al. (2021)).