Page 318 - MEGIN Book Of Abstracts - 2023

In contrast to the control participants, rs-MEG source beta band yielded an accuracy of 94.19% and a sensi-
imaging in the children with mTBI showed: 1) hyperac- tivity of 90.00%, when tested with our mTBI dataset.
tivity from the bilateral insular cortices in alpha, beta, The results support the suggestion of multi-instance
and low-frequency bands, from the left amygdala in one-class SVM for the detection of mTBI.
alpha band, and from the left precuneus in beta band;
2) hypoactivity from the bilateral dorsolateral prefrontal IEEE transactions on neural systems and rehabilitation
cortices (dlPFC) in alpha and beta bands, from the ven- engineering: a publication of the IEEE Engineering
tromedial prefrontal cortex (vmPFC) in beta band, from in Medicine and Biology Society (2020), Vol. 28, No. 1
the ventrolateral prefrontal cortex (vlPFC) in gamma (31647439) (4 citations)
band, from the anterior cingulate cortex (ACC) in alpha
band, and from the right precuneus in alpha band.
The present study showed that MEG source imaging Marked Increases in Resting-State MEG Gamma-
technique revealed abnormalities in the resting-state Band Activity in Combat-Related Mild Traumatic
electromagnetic signals from the children with mTBI. Brain Injury (2020)

Keywords: MEG, mild traumatic brain injury, pediatric, Huang, Ming-Xiong; Huang, Charles W; Harrington,
post-concussion symptoms, psychiatric disorder Deborah L; Nichols, Sharon; Robb-Swan, Ashley;
Angeles-Quinto, Annemarie; Le, Lu; Rimmele, Carl;
Journal of neurotrauma (2020), Vol. 37, No. 7 (31724480) (5 Drake, Angela; Song, Tao; Huang, Jeffrey W; Clifford,
citations) Royce; Ji, Zhengwei; Cheng, Chung-Kuan; Lerman,
Imanuel; Yurgil, Kate A; Lee, Roland R; Baker, Dewleen G

Anomaly Detection of Moderate Traumatic Brain Department of Radiology, University of California, San Diego,
Injury Using Auto-Regularized Multi-Instance One- CA, USA; Department of Bioengineering, Stanford University,
Class SVM (2020) Stanford, CA, USA; Department of Neuroscience, University
of California, San Diego, CA, USA; ASPIRE Center, VASDHS
Rasheed, Waqas; Tang, Tong Boon Residential Rehabilitation Treatment Program, San Diego, CA,
USA; Cedar Sinai Medical Group Chronic Pain Program, Bev-
ABSTRACT Detection and quantification of func- erly Hills, CA, USA; Department of Computer Science, Colum-
tional deficits due to moderate traumatic brain injury bia University, New York, NY, USA; VA Center of Excellence for
(mTBI) is crucial for clinical decision-making and timely Stress and Mental Health, San Diego, CA, USA; Department of
commencement of functional therapy. In this work, Computer Science and Engineering, University of California,
we explore magnetoencephalography (MEG) based San Diego, CA, USA; Radiology, Research, and Psychiatry Ser-
functional connectivity features i.e. magnitude squared vices, VA San Diego Healthcare System, San Diego, CA, USA;
coherence (MSC) and phase lag index (PLI) to quantify Department of Psychological Sciences, Loyola University,
synchronized brain activity patterns as a means to New Orleans, LA, USA
detect functional deficits. We propose a multi-instance
one-class support vector machine (SVM) model gener- ABSTRACT Combat-related mild traumatic brain injury
ated from a healthy control population. Any disper- (mTBI) is a leading cause of sustained impairments in
sion from the decision boundary of the model would military service members and veterans. Recent animal
be identified as an anomaly instance of mTBI case studies show that GABA-ergic parvalbumin-positive in-
(Glasgow Coma Scale, GCS score between 9 and 13). terneurons are susceptible to brain injury, with damage
The decision boundary was optimized by considering causing abnormal increases in spontaneous gamma-
the closest anomaly (GCS =13) from the negative class band (30-80 Hz) activity. We investigated spontaneous
as a support vector. Validated against magnetic reso- gamma activity in individuals with mTBI using high-
nance imaging (MRI) data, the proposed model at high resolution resting-state magnetoencephalography

 ontents  Index 297
C