In this chapter, we provide an automated computational algorithm for detection of traumatic brain injury (TBI) from T2-weighted magnetic resonance (MRI) images. The algorithm uses a combination of brain symmetry and 3D connectivity in order to detect the regions of injury. The images are preprocessed by removing all non-brain tissue components. The ability of our symmetry-based algorithm to detect the TBI lesion is compared to manual detection. While manual detection is very operator-dependent which can introduce intra-and inter-operator error, the automated detection method does not have these limitations and can perform skull stripping and lesion detection in real-time and more rapidly than manual detection. The symmetry-based algorithm was able to detect the lesion in all TBI animal groups with no false positives when it was tested versus a naive animal control group.
CITATION STYLE
Esfahani, E. T., McBride, D. W., Shafiei, S. B., & Obenaus, A. (2015). A Real-Time Analysis of Traumatic Brain Injury from T2 Weighted Magnetic Resonance Images Using a Symmetry-Based Algorithm (pp. 99–117). https://doi.org/10.1007/978-3-319-23724-4_5
Mendeley helps you to discover research relevant for your work.