BrainGNN

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Morbi nec metus justo. Aliquam erat volutpat.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Morbi nec metus justo. Aliquam erat volutpat.

• 23 TLE, 32 Normal Controls
• Accuracy of 94.5%, .99 AUC

Details:

• 8.5 minute of scanning
  • 3.4*3.4*3.5 mm voxel
  • Smoothed with 14 mm FWHM
• Used two atlases: Dosenbach-160 and AAL-90??
• Thresholding done using Global Cost Efficiency?
• From binary graph calculated:
  • Flow coefficient
  • k-core centrality
  • pagerank centrality
  • betweeness centrality
  • (g) clustering coefficient
  • (g) global efficiency
  • (g) local efficiency
  • (g) node shortest path
• Then each feature, used two sample t-test
• Then SVM

Title Text

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Morbi nec metus justo. Aliquam erat volutpat.