• Ahmad Alhourani, Sudhir K Pathak, Michael J Randazzo, Tom Wozny, Efstathios Kondylis, Shawn Walls, Michael Ward, Stephen Foldes, Donald Krieger, David O Okonkwo, R Mark Richardson, and Ajay Niranjan.
• Neurosurgery. 2015 Aug 1;62 Suppl 1:227.

IntroductionConcussion remains a clinical diagnosis with the lack of objective changes on standard brain imaging. Failure to document concussion delays appropriate intervention in this at-risk population. Magnetoencephalography (MEG) is a powerful, noninvasive imaging modality which may offer unique insight into functional brain networks affected in concussed patients.MethodsTen symptomatic concussion patients with normal MR imaging and 7 healthy controls underwent a 5 minute MEG recording during quiet rest using a whole-head 306 channel Neuromag Vectorview System. Scans were done within a few weeks to months of injury while still symptomatic. Analysis was performed using MNE python suite and Matlab software. The strength of functional connections between cortical regions was measured using the phase locking value (PLV) between all individual sensors. PLV was computed within the delta (1-4 Hz), theta (4-8 Hz), low alpha (8-10 Hz), and high alpha (10-13 Hz) frequency bands. The small world index (SWI) was calculated from the adjacency matrix of pairwise PLV. An SWI value of 1 represents an ideal network that is balanced in terms of local connectivity and global integration. Deviations below 1 signifies a change to a more random network.ResultsThe SWI in concussion subjects was significantly lower than healthy controls in the delta (concussion mean = 0.8777, controls mean = 0.9067, P = .0458) and high alpha frequency bands (concussion mean = 0.8793, controls mean = 0.9199, P = .0232). This lower SWI score indicates an increase in network randomness, suggesting less efficient communication selectively in the delta and high alpha bands.ConclusionThese results suggest that short-duration MEG recording may be used to diagnose concussion. MEG can be used to calculate an objective measure quantifying the degree of global dysfunction that potentially can be used to predict functional outcomes or symptom severity. This is the first study to apply graph theory on MEG-acquired data in concussion.

11 keywords...

hide…