Meeting Banner
Abstract #2688

Detection of Mild Traumatic Brain Injury Utilizing Multifeature Analysis of MRI

Yongxia Zhou1, Yao Wang2, Damon Kenul3, Yuanyi Xue2, Yulin Ge3, Joseph Reaume3, Robert I. Grossman4, Yvonne W. Lui3

1Radiology/Center for Biomedical Imaging, New York University Langone Medical Center, New York, NY, United States; 2Electrical & Computer Engineering, Polytechnic Institute of New York University, Brooklyn, NY, United States; 3Radiology/Center for Biomedical Engineering, New York University Langone Medical Center, New York, NY, United States; 4Radiology/Center for Biomedical Engineering, New York University, New York, NY, United States

The purpose of this study is to design and develop computational techniques to identify mild traumatic brain injury (MTBI) patients that can be used to help predict patient long-term outcome ultimately using multi-dimensional feature space based on several advanced quantitative MR measures. Fourteen imaging features (e.g. kurtosis, magnetic field correlations, thalamic network connectivity and regional volumetry), and nineteen clinical features were tested with different feature selection and classifier algorithms. Our study demonstrates that an automatic classification based on objective physical and imaging measures can achieve a high accuracy of nearly 100% and a robust prediction for the long-term outcome (P0.01).

Abstract PDF

Keywords

accuracy accurate achieve acquisition activation analyzed anterior anxiety averaging axial best biomedical brain bring calculating central cerebrum classification classifier classifiers classifying clinical coil combining component contribute control controls correlation cortical criteria cross days defense definitions detection diffusion disease electrical engineering entire evaluated excitation eyes facilitate fatigue feature features female field findings fold frontal functional gradient gray health healthy helpful hours house identify improved include initial injury international interval kurtosis layer linear long macroscopic males maps matrix medical medicine metrics microscopic mild minutes namely network none novel objective organizations original outcome patients pilot planar positioned practice predicting prediction primary processed pulse radiology refocusing regional registration regression regressors resolution resting rule score scripts select selection sense sensitivity several significantly signs slice slices software solutions spatial specifically specificity spin statistical stay structures studied subject subjects symptom term thalamus threshold training traumatic turbo understanding useful utilizing validation verbal visit volume volumes white whole year years