A two-level multimodality imaging Bayesian network approach for classification of partial epilepsy: preliminary data

Neuroimage. 2013 May 1;71:224-32. doi: 10.1016/j.neuroimage.2013.01.014. Epub 2013 Jan 24.

Abstract

Background: Quantitative neuroimaging analyses have demonstrated gray and white matter abnormalities in group comparisons of different types of non-lesional partial epilepsy. It is unknown to what degree these type-specific patterns exist in individual patients and if they could be exploited for diagnostic purposes. In this study, a two-level multi-modality imaging Bayesian network approach is proposed that uses information about individual gray matter volume loss and white matter integrity to classify non-lesional temporal lobe epilepsy with (TLE-MTS) and without (TLE-no) mesial-temporal sclerosis and frontal lobe epilepsy (FLE).

Methods: 25 controls, 19 TLE-MTS, 22 TLE-no and 14 FLE were studied on a 4T MRI and T1 weighted structural and DTI images acquired. Spatially normalized gray matter (GM) and fractional anisotropy (FA) abnormality maps (binary maps with voxels 1 SD below control mean) were calculated for each subject. At the first level, each group's abnormality maps were compared with those from all the other groups using Graphical-Model-based Morphometric Analysis (GAMMA). GAMMA uses a Bayesian network and a Markov random field based contextual clustering method to produce maps of voxels that provide the maximal distinction between two groups and calculates a probability distribution and a group assignment based on this information. The information was then combined in a second level Bayesian network and the probability of each subject to belong to one of the three epilepsy types calculated.

Results: The specificities of the two level Bayesian network to distinguish between the three patient groups were 0.87 for TLE-MTS and TLE-no and 0.86 for FLE, the corresponding sensitivities were 0.84 for TLE-MTS, 0.72 for TLE-no and 0.64 for FLE.

Conclusion: The two-level multi-modality Bayesian network approach was able to distinguish between the three epilepsy types with a reasonably high accuracy even though the majority of the images were completely normal on visual inspection.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Adult
  • Bayes Theorem
  • Brain Mapping*
  • Diffusion Tensor Imaging
  • Epilepsies, Partial / classification*
  • Female
  • Humans
  • Image Interpretation, Computer-Assisted / methods*
  • Male