A computational model of epidural electrical stimulation of the cervical spinal cord in non-human primates

Abstract: Cervical spinal cord injury alters the communication between the brain and the spinal circuits controlling movement, often leading to tetraplegia. Epidural Electrical Stimulation (EES) of the lumbar spinal cord has shown promising results to restore leg motor control after paralysis. EES modulates the activity of proprioceptive afferent circuits, enabling the spinal cord to elaborate coordinated movements of previously paralyzed limbs. Similar proprioceptive afferent circuits contribute to upper-limb motor control, suggesting that EES may also improve the recovery of upper-limb movements after injury. The ability to engage individual or small groups of muscles is essential to facilitate motor control with EES. At this stage, however, this ability remains largely unexplored.
To address this question, we developed a realistic Finite Element/axon-cable biophysical model of EES applied to the non-human primate cervical spinal cord. Our objective was to evaluate and optimize the specificity of tailored, dura mater-like electrode implants placed dorsally over the spinal cord. The anatomically realistic model was derived from CT-scan acquisitions that supported 3D-reconstruction of the cervical vertebrae. We inserted physical compartments for the electrode silicone paddle and for the spinal roots, and used curvilinear coordinates to represent the white matter and spinal roots conductivity anisotropy.
We used the model to quantify the recruitment of Group I and Group II afferent fibers in the dorsal roots, motor axons in the ventral roots, and large myelinated fibers in the dorsal columns. To validate our model, we estimated the muscle responses to single pulses of EES using a realistic connectivity model between Ia-afferents and motoneurons innervating upper-limb muscles. We then compared our results to experimental recordings performed in two macaque monkeys under anesthesia.
We found a high correlation between the responses derived from simulations and obtained in vivo. However, the anatomical features exerted a non-negligible impact on the predicted recruitments. These results emphasize the importance of including realistic anatomical features to derive implant specificity from computer simulations.

Finally, we found that lateralized epidural stimulation of the cervical spinal cord recruits individual dorsal roots at significantly lower thresholds than other neighboring structures, suggesting that targeted EES could selectively modulate upper-limb motor pools. Taken together, these results establish the framework for the design of targeted cervical implants to facilitate upper-limb movements after spinal cord injury.

1Ctr. for Neuroprosthetics, Brain Mind Inst., EPFL, Geneve, Switzerland; 2Domain of Neurophysiology, Dept. of Med., Univ. of Fribourg, Fribourg, Switzerland; 3Ctr. Hospitalier Universitaire Vaudois (CHUV), Lausanne, Switzerland
N. Greiner: None. B. Barra: None. S. Borgognon: None. G. Schiavone: None. S. Lacour: None. J. Bloch: None. E.M. Rouiller: None. G. Courtine: None. M. Capogrosso: None.

This entry was posted in Biomaterials, Chronic Spinal Cord Injury Research, Neuroscience Abstracts, Rehabilitation, spinal cord injury research. Bookmark the permalink.