Poster AA4 - Ultrasound Echogenicity Correlates with Dorsiflexion Torque During Noninvasive Spinal Stimulation with Functional Electrical Stimulation

Introduction: Spinal cord stimulation (SCS) and functional electrical stimulation (FES) are two distinct approaches to rehabilitate people with spinal cord injury. FES activates skeletal muscles directly through direct stimulation of peripheral nerves and improves muscle and bone health. In contrast, SCS can excite the peripheral motor pathways of the skeletal muscles via spinal circuitry. FES usually generates high-amplitude muscle force but is susceptible to muscle fatigue. SCS-elicited muscle force varies and is unpredictable, but as SCS physiologically recruits the muscle fibers, it is less susceptible to muscle fatigue. Evidence indicates that combining these modalities may produce greater muscle activation than either method alone; however, measurements of muscle activation have been largely limited to electromyography, which becomes contaminated by stimulation artifacts during continuous FES or SCS. Alternatively, ultrasound is an emerging technique to quantify volitionally or FES-evoked muscle activation. However, its use to detect muscle activation evoked via SCS has not been shown. In this preliminary experiment, we demonstrate that ultrasound (US) echogenicity correlates with both FES-evoked and SCS-generated limb force and also detects the increased force output during combined stimulation (FES and SCS) of the ankle dorsiflexors.

Materials and

Methods: One participant with no disabilities completed the task described herein. One 3.2cm electrode was placed midline between the T11 and T12 spinous process. Two 7.5x10cm return electrodes were placed symmetrically on the abdomen. For FES, two 5x5cm electrodes were placed on the ankle dorsiflexors. The subject's foot was secured in a testbed with a load cell to ensure isometric contractions. All stimuli were supplied as triplets (frequency =16.67 Hz, pulsewidth = 0.5ms). An EMG sensor was secured on the subject's TA to verify motor activation.
First, tests were performed to identify the subject's saturation levels. For FES, three triplets were supplied from 5 mA to 49 mA in intervals of 2 mA. For SCS, three triplets were supplied from 30 mA to 80 mA in intervals of 5 mA. For subsequent tests, SCS amplitude was set to 35 mA (53.85% saturation), and FES amplitude was set to 22.5 mA (50% saturation).
Tests were performed with SCS only, FES only, and SCS with FES. In each case, three sets of triplets were applied. In the SCS + FES case, the relative stimulation interval (RSI) between SCS onset and FES onset varied from 0.105 to 0.145ms. A US probe was placed over the dorsiflexors, and force and US data were collected during each stimulus at a rate of 1 kHz. The normalized peak force and normalized US echogenicity were compared across cases. Normalized force integral was also compared for the cases involving FES.

Results, Conclusions, and Discussions: The application of SCS prior to FES increased average normalized peak force by 8.84% (mean + SD: FES only: 0.87 + 0.02, FES + SCS: 0.94 + 0.07). Similarly, applying SCS prior to FES increased average normalized force integral (from FES onset to 1s after onset) by 14.52% (FES only: 0.76 + 0.03, FES + SCS: 0.88 + 0.12). Increase in peak force was greatest for the longest RSI (0.145 ms), whereas increase in force integral was greatest for the shortest RSI (0.105 ms). As expected, average normalized peak force for FES only and FES + SCS far exceeded that of SCS only. The FES only and FES + SCS cases exhibited peak force outputs equal to 371.53% and 412.11% that of the SCS case, respectively.
Note that for cases with FES, the trigger signal for US data collection was sent to the US imaging machine 0.292s before the trigger signal was sent to the laptop controlling the FES stimulator, and US data collection terminated after 0.750s. This is reflected in the echogenicity plots. For cases with SCS, the SCS trigger preceded the US trigger by 0.187 and 0.167s for the first and second triplets , respectively. US was not collected for the third triplet due to an error in US triggering.
For each 2-D US image, mean echogenicity was calculated across the entire image. The data were then normalized to the maximum and minimum echogenicity values collected across all conditions. To account for differences in initial echogenicity value, the maximum decrease in echogenicity from its initial value was calculated for each 0.750s imaging sequence. Mean maximum decrease in echogenicity was 48.21% greater for FES + SCS (0.81 + 0.04) than for FES only (0.55 + 0.07).
These results are in agreement with existing literature demonstrating that application of SCS prior to FES increases force output. Furthermore, they illustrate that increased echogenicity decay reflects this increased force. Ultrasound may therefore offer a tool for measuring muscle activation during tasks involving continuous SCS with FES without the need for filtering of stimulation artifacts.

Acknowledgements (Optional):