Multiple sclerosis cortical lesion detection with deep learning at ultra-high-field MRI

This is the code repository of the NMR in Biomedicine paper.

Overview

This software provide a cortical lesion (CL) segmentation and classification given 7T MP2RAGE images. The code is build upon MONAI and it is based on Pytorch. A GPU is required to train the network with new data, whereas a LINUX machine is sufficient for inference.

Install

Run

License

This software is released under the version 2.0 of the Apache License.

Cite

If you use this code please cite the following references:

• La Rosa, Francesco, et al. "Multiple sclerosis cortical lesion detection with deep learning at ultra‐high‐field MRI." NMR in Biomedicine (2022): e4730.. doi
• La Rosa, Francesco, et al. "Automated detection of cortical lesions in multiple sclerosis patients with 7t mri." International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, Cham, 2020. doi
• The MONAI Consortium. (2020). Project MONAI. Zenodo. doi