Spinal Nerve Modeling Within The Spinal Canal

Abstract

The objective of this work is to develop a high‐resolution model of the brachial plexus with detail including individual spinal nerve rootlets. Medical illustrations may include this level of detail with artistic enhancement of structures of interest and suppression of background visual noise. Careful cadaver dissection of an open spinal canal and separated dural sheath will reveal the detail we seek for our digital model, but such dissection usually demonstrates them in positions unlike that of the living and intact nervous system. Our goal is to segment the nerves forming the brachial plexus, retaining their positions in‐vivo. We use a high resolution Visible Human Project type block face image dataset. The image database is from whole‐body imaging of an elderly, female donor with visible pathology. The focus of this work was on modeling the spinal nerves from C3 to T1. The brachial plexus and some peripheral nerves were included for orientation. This image database is made up of voxels spaced orthogonally 63μ apart, compared to the Visible Human Male image database with voxels spaced 1000μ apart. This voxel size yields a spatial resolution on the order of 300μ for this new dataset compared to the 5mm resolution of the VH Male. We can now see individual rootlets exiting the spinal cord to form the anterior and posterior spinal nerve roots but it is still challenging to follow each rootlet from beginning to end. We concentrated on nerves C3 to T1 which are the most visible as they diverge from the spinal cord soon after exiting the spinal columns. We were able to identify an average of five rootlets for each anterior and posterior root for nerves C3 through T1. Splines were constructed with control points to fit the path of the nervous structures. Each control point was assigned a radius to approximate the size of the structure at that point. We identified an anatomic anomaly demonstrating a conjoined posterior rootlet, which emerged from the posterior gray horn as one rootlet but split to contribute to the roots of C6 and C7. Within the brachial plexus we were able to identify the posterior rami for spinal nerves C3 to C5, which we were unable to find in VH Male data. For nerves C5 through C8 we followed anterior rami into the trunks, divisions, and cords of the brachial plexus. Following divisions and cords proved difficult because the cadaver was oriented with arms at her side such that the brachial plexus was compact as it is in the anatomical position. Freezing artifacts occasionally obscured the tissue, making it hard to confidently follow specific nerve components in a cluster of multiple trunks or divisions. We were still able to identify the superior, middle, and inferior trunks of the brachial plexus, the posterior and anterior divisions of each trunk, the lateral, posterior, and medial cords, the terminal nerves, and branching peripheral nerves. This modeling of the cervical spinal nerves and the brachial plexus is not only unique because of the extraordinary detail of the nervous structures, but also advantageous because it allows anatomy students to visualize these nerves in the same spatial configuration as in the living human.Figure shows 3D splines of brachial plexus and spinal nerve rami, roots, and and rootlets of C3 through T1.Figure 1Figure shows anatomical anomaly found where two nerve rootlets coming out from the root of C6 and C7 conjoin into a single rootlet before entering the posterior gray horn.Figure 2