eduzhai > Applied Sciences > Engineering >

Automated Detection of Cortical Lesions in Multiple Sclerosis Patients with 7T MRI

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 11 pages

Abstract: The automated detection of cortical lesions (CLs) in patients with multiplesclerosis (MS) is a challenging task that, despite its clinical relevance, hasreceived very little attention. Accurate detection of the small and scarcelesions requires specialized sequences and high or ultra-high field MRI. Forsupervised training based on multimodal structural MRI at 7T, two expertsgenerated ground truth segmentation masks of 60 patients with 2014 CLs. Weimplemented a simplified 3D U-Net with three resolution levels (3D U-Net-). Byincreasing the complexity of the task (adding brain tissue segmentation), whilerandomly dropping input channels during training, we improved the performancecompared to the baseline. Considering a minimum lesion size of 0.75 { mu}L, weachieved a lesion-wise cortical lesion detection rate of 67 and a falsepositive rate of 42 . However, 393 (24 ) of the lesions reported as falsepositives were post-hoc confirmed as potential or definite lesions by anexpert. This indicates the potential of the proposed method to support expertsin the tedious process of CL manual segmentation.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×