Building a tractography model involves complex, multi-step processes and requires continuous decision-making under uncertainty at various stages. As a PhD student, my focus is on developing multi-task, self-interpretable transformers to build a probabilistic tractography model. I aim to train this model for various tasks, including denoising, segmentation, diffusion reconstruction, and fiber tracking. Existing dMRI datasets will be used to train the model and our agent will be able to process multi-modal data.
DRCMR:
Nordre Hospitalsvej 13 2650 Hvidovre Captial Region of Denmark
DRCMR Main WebsiteContact: timd@drcmr.dk
DTU Compute:
Bygning 324 2800 Kongens Lyngby Captial Region of Denmark
DTU Compute Main WebsiteContact: tbdy@dtu.dk
