Extended Data Fig. 9: Neuroprosthetic baroreflex implemented in non-human primates with spinal cord injury.
From: Neuroprosthetic baroreflex controls haemodynamics after spinal cord injury
Step 1(a): We recorded blood pressure using an axillary artery catheterization preparation, and calculated the peak change in blood pressure during stimulation. We found that there was a haemodynamic hotspot in the lower thoracic spinal cord, specifically over segments T10, T11 and T12. Step 2(b): Accounting for the offset in segments between species, we found a linear relationship between the functional mapping results in rats and the functional mapping results in non-human primates, further confirming the localization of a haemodynamic hotspot in the last three thoracic segments (linear modelling; R2 = 0.65; P = 0.018). Step 3(c): To develop the spatial features of an electrode array specifically targeting haemodynamic hotspots in the non-human primate we quantified all the features of the low thoracic spinal column. We measured the precise length of each spinal segment and vertebra using a combination of gross anatomical dissections, high resolution computed tomography scans, and custom MRI sequences. Step 4(d): We used these anatomical features, driven by the identification of haemodynamic hotspots to develop an electronic dura mater specifically targeting T10, T11 and T12 spinal segments. Detailed dimensions of the active stimulation sites (0.7 mm diameter, 0.014 cm2 geometric surface area). Step 5(e): Electrochemical impedance spectrum (modulus, left and phase, right) of an electrode array acquired in vitro post-fabrication indicates functional, low-impedance electrodes. Step 6(f): We fully implemented a negative-pressure chamber designed to fit a non-human primate to induce a stimulated orthostatic challenge. We found that stimulation using the chamber led to an immediate and consistent decrease in blood pressure. Step 7(g): We next implemented our biomimetic stimulation protocols. We found a linear relationship between stimulation amplitude and the pressor response (n = 3 monkeys; linear modelling, R2 = 0.85; P = 9.33 × 10−5). Compared to stimulating only one hotspot, stimulating with all three sets of electrodes, and therefore targeting each haemodynamic hotspot led to a more robust increase in blood pressure (n = 2 monkeys). Step 8(h): Implementation of the neuroprosthetic baroreflex, acting in closed loop, led to sustained increases in blood pressure that did not fatigue (n = 1 monkey). In comparison, continuous open-loop stimulation, using the same stimulation parameters, elicited an increase in blood pressure that was immediately followed by rapid fatigue. For continuous stimulation, the amplitude was set to the maximum observed stimulation value in the trials with the closed-loop controller. Only biomimetic stimulation re-established the natural frequency dynamics, revealed using wavelet decomposition (n = 2 monkeys). Step 9(i): Using this closed loop approach, we found a reduction in the target (baseline) error, reduced time outside key thresholds (−5 mmHg; −2 mmHg for diastolic blood pressure), a restoration of the nonlinear relationship between blood pressure and chamber pressure, and a concomitant reduction in the linear model coefficient. These quantifications held for systolic blood pressure (top), diastolic blood pressure (middle), and mean arterial pressure (bottom).
