Meeting Banner
Abstract #1318

Adapting a white matter lesion segmentation algorithm for large cohort studies

Leonie Lampe 1,2 , Alexander Schaefer 1,3 , Christopher J. Steele 1 , Katrin Arlin 1,2 , Dominik Fritzsch 4 , Matthias L. Schroeter 1,2 , Arno Villringer 1,2 , and Pierre-Louis Bazin 1

1 Department of Neurology, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, 2 Leipzig Research Centre for Civilization Diseases & Clinic of Cognitive Neurology, University of Leipzig, Germany, 3 Clinical Imaging Research Centre & Singapore Institute for Neurotechnology, National University of Singapore, Singapore, 4 Department of Neuroradiology, University Hospital Leipzig, Germany

Here we adapted and validated a lesion segmentation algorithm previously aimed at MS lesions for white matter lesions (WML) segmentation within the general population. WML in the normal aging brain display diversity in pattern, intensity and extent. By means of iteratively re-normalising the contrast of the FLAIR images to better separate lesions from healthy tissue a dice coefficient of 0.63 was obtained. The validation was performed with 5 subjects with diverse lesions. The algorithm was applied to a large cohort study (age range 19-80 years) with approximately 1200 subjects.

This abstract and the presentation materials are available to members only; a login is required.

Join Here