A preliminary exploration of brain fiber bundle reconstruction and its influencing factors based on diffusion tensor imaging

Abstract

This paper focuses on the methods and progress of brain fiber bundle construction based on diffusion tensor imaging data, and conducts a preliminary exploration of the construction results. The study is based on 45 subjects in the public database, using different software to construct nerve fiber bundles for comparison, calculating the number of fiber bundle tracking under different parameters, and finally conducting a t-test on the fiber bundle connection strength of the left and right brain. The nerve fibers constructed by different software are relatively consistent and also consistent with human brain nerve fibers. A more relaxed length threshold will increase the number of fiber tracking, and as the angle threshold increases, the number of fiber tracking will first increase and then decrease. There are differences in the connection strength of the left and right brain. This preliminary study provides a foundation for further exploring brain connectivity and analyzing differences between individuals or patient groups. While the fiber bundles constructed matched expectations, more validation is needed against post-mortem dissections or other imaging techniques. Additionally, expanding the study to include more subjects and analyzing specific fiber bundles or regions could provide more detailed insights. The methods and parameters evaluated in this work help establish best practices for fiber tracking and analyzing connectivity strength metrics. With refinement, such fiber bundle analysis tools may help researchers better understand structural brain networks and differences related to development, aging or disease.