Brain White Matter Network Measures for Cerebral Palsy

作者：Li, Jun; Wang, Yuanjun^*; Yang, Cheng

来源：Journal of Medical Imaging and Health Informatics, 2018, 8(3): 419-424.

DOI：10.1166/jmihi.2018.2330

• Cerebral Palsy
• Diffusion Tensor Image
• White Matter Network
• Graph Theory

Purpose: To investigate the whole brain white matter (WM) network and disrupted topological organization in network of cerebral palsy (CP) infants. Materials and Methods: Diffusion tensor images are obtained from 15 CP infants and 30 gender and age matched normal infants. To define the WM network, Pediatric Brain Atlas 83(PBA 83) is used for identifying 83 nodes, and Fiber Assignment by Continuous Tracking (FACT) tractography algorithm is employed for nerve fiber construction. Finally graph theory and network based statistic approach are used for the WM network analysis. Results: Both CP and normal infants WM network are small world topology (sigma >> 1). However, comparing with normal infants, CP infants have significantly decreased clustering coefficient (p = 0.0055), global efficiency (p = 0.0217) and local efficiency (p = 0.0257), but increased shortest path length (p = 0.0382) in WM networks. In addition, many nodes exhibit significantly decreased nodal centralities (p < 0.05) involving nodal degree, efficiency and nodal clustering coefficient, and these abnormal regions mainly involve temporal lobe, frontal lobe, parietal lobe, the insula and cingulate gyri and occipital lobe. Conclusion: CP infants have significantly altered WM network parameters and regional nodal characteristics, implying the WM lost connectivity and topological disorganization. The result is consistent with previous studies, but more accurate in the specific cerebral cortex. And the WM network analysis based on a priori atlas can provide a novel choice to quantify the CP WM lesions and study CP brain connectivity lost.

• 出版日期2018-3
• 单位上海理工大学