Research data supporting "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy"
Repository URI
Repository DOI
No Thumbnail Available
Type
Dataset
Change log
Authors
Adler, S
Wagstyl, K
Gunny, R
Carmichael, D
Description
Data used to train neural network classifier for FCD detection in a paediatric cohort. These are surface-based cortical features derived from structural MRI data, such as cortical thickness, FLAIR intensity etc. The scripts to generate the surface based features are available through freesurfer (https://surfer.nmr.mgh.harvard.edu/) and our github (https://github.com/kwagstyl/FCDdetection).
Version
Software / Usage instructions
Input data were generated using freesurfer and matlab scripts found on https://github.com/kwagstyl/FCDdetection. The main .mat file contains a matrix of patients' features as described in the accompanying .csv file. Data are registered to the left hemisphere of fsaverage_sym - the bilaterally symmetrical template subject - and only include cortical vertices (lh.cortex label). The correct classification of vertices into lesional and non-lesion is also included as a column. The intention is to provide training data with which different classifiers could be tested.
Keywords
FCD, Machine learning classifier, FreeSurfer, structural MRI
