• DocumentCode
    2156409
  • Title

    Feature analysis of functional MRI for discrimination between normal and epileptogenic brain

  • Author

    Burrell, Lauren S. ; Glynn, Simon M. ; Vachtsevanos, George J. ; Litt, Brian

  • Author_Institution
    Georgia Inst. of Technol., Atlanta, GA, USA
  • fYear
    2007
  • fDate
    2-5 July 2007
  • Firstpage
    2554
  • Lastpage
    2559
  • Abstract
    Temporal lobe epilepsy (TLE) is the most common epilepsy syndrome in adults. As it is a focal disorder, i.e. limited to a particular area of the brain, it is often curable through surgery when drug therapies fail to control seizures. In order to ensure successful surgical outcome, the abnormal brain regions must be localized as accurately as possible to prevent the removal of healthy tissue. We report on a novel voxel-based procedure for discrimination between normal and epileptogenic brain tissue through feature analysis of continuous arterial spin labeling (ASL) perfusion functional magnetic resonance imaging (fMRI) data. Five TLE patients and three healthy controls were studied. Features were extracted from the fMRI time series of each subject to determine which individual features and combinations of features could correctly separate epileptogenic and normal brain tissue.
  • Keywords
    biological tissues; biomedical MRI; brain; feature extraction; geriatrics; medical disorders; medical image processing; neurophysiology; surgery; TLE patients; abnormal brain regions; adults; continuous arterial spin labeling perfusion; drug therapy; epilepsy syndrome; epileptogenic brain tissue; fMRI data; feature analysis; focal disorder; functional magnetic resonance imaging data; normal brain tissue; surgery; temporal lobe epilepsy; time series; voxel-based procedure; Accuracy; Blood; Electroencephalography; Epilepsy; Feature extraction; Hippocampus; Feature extraction; functional magnetic resonance imaging (fMRI); genetic programming;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Control Conference (ECC), 2007 European
  • Conference_Location
    Kos
  • Print_ISBN
    978-3-9524173-8-6
  • Type

    conf

  • Filename
    7068386