Department of Biomedical Engineering Final Oral Examination

Zachariah Sperry

Neural Interfacing with Dorsal Root Ganglia: Anatomical Characterization and Electrophysiological Recordings with Novel Electrode Arrays

Dorsal root ganglia (DRG), the hubs of neurons conducting sensory information into the spinal cord, are promising targets for clinical and investigative neural interface technologies. DRG stimulation is currently a tertiary therapy for chronic pain patients, which has an estimated prevalence of up to 40% of adults in the United States. In pre-clinical studies, combined neural recording and stimulation at DRG has been used as part of novel closed-loop systems to drive activity of the limbs or the urinary system. This suggests a potential role for clinical DRG interfaces to assist, among other patient groups, the nearly 300,000 spinal cord injured patients in the United States.

To maximize the utility of DRG interfaces, however, there remains a strong need to improve our understanding of DRG structure. Neural interface technologies for both stimulation and recording rely heavily on the spatial organization of their neural targets. To record high-fidelity neural signals, a microelectrode must be placed within about 200 µm of a neural cell body. Likewise, effective neural stimulation is believed to act on a subset of DRG axons based on their size and location. The spatial organization of DRG, however, has not been previously quantified. In this thesis, I demonstrate a novel algorithm to transform histological cross-sections of DRG to a normalized circular region for quantifying trends across many samples. Using this algorithm on 26 lumbosacral DRG from felines, a common preclinical DRG model, I found that the highest density of neural cell bodies was in the outer 24% radially, primarily at the dorsal aspect. I extended this analysis to a semi-automated cross-DRG analysis in 33 lower lumbar DRG from 10 human donors. I found that the organization of human DRG was similar to felines, with the highest density of cell bodies found in the outer 20-25% of the DRG, depending on spinal level. I also found a trend toward lower small-axon density at the dorsal aspect of L5 DRG, a key region for stimulation applications.

To take advantage of this quantitative knowledge of DRG organization, future neural interfaces with DRG will require more advanced technologies. Standard silicon-based electrode arrays, while useful for short-term DRG recordings, ultimately fail in chronic use after several weeks as a result of mechanical mismatch with neural tissue. In this thesis, I demonstrate sensory recording from the surface and interior of sacral DRG during acute surgeries using a variety of flexible polyimide microelectrode arrays 4 μm thick with minimum site separation of 25 to 40 μm. Using these arrays, I recorded from bladder and somatic afferents with high fidelity. The high density of sites allowed for neural source localization from surface recordings to depths 25 to 107 µm. This finding supports the anatomical analysis suggesting a high density of cell bodies in the dorsal surface region where the surface array was applied. The high site density also allowed for the use of advanced signal processing to decrease analysis time and track neural sources during movement of the array which may occur during long-term behavioral experiments.

This thesis represents significant advances in our understanding of DRG and how to interface with them, particularly related to the ways anatomy can inform development of future technologies. Going forward, it will be important to expand the anatomical maps based on organ function and to test the novel flexible arrays in chronic implant experiments.

Date: Wednesday, September 11, 2019
Time: 9:00 AM
Location: NCRC Building 32 Auditorium
Chair: Dr. Tim Bruns