Motorless cadence control of standard and low duty cycle-patterned neural stimulation intensity extends muscle-driven cycling output after paralysis

Abstract

Background: Stimulation-driven exercise is often limited by rapid fatigue of the activated muscles. Selective neural stimulation patterns that decrease activated fiber overlap and/or duty cycle improve cycling exercise duration and intensity. However, unequal outputs from independently activated fiber populations may cause large discrepancies in power production and crank angle velocity among pedal revolutions. Enforcing a constant cadence through feedback control of stimulus levels may address this issue and further improve endurance by targeting a submaximal but higher than steady-state exercise intensity.

Methods: Seven participants with paralysis cycled using standard cadence-controlled stimulation (S-Cont). Four of those participants also cycled with a low duty cycle (carousel) cadence-controlled stimulation scheme (C-Cont). S-Cont and C-Cont patterns were compared with conventional maximal stimulation (S-Max). Outcome measures include total work (W), end power (P_end), power fluctuation (PFI), charge accumulation (Q) and efficiency (η). Physiological measurements of muscle oxygenation (SmO₂) and heart rate were also collected with select participants.

Results: At least one cadence-controlled stimulation pattern (S-Cont or C-Cont) improved P_end over S-Max in all participants and increased W in three participants. Both controlled patterns increased Q and η and reduced PFI compared with S-Max and prior open-loop studies. S-Cont stimulation also delayed declines in SmO2 and increased heart rate in one participant compared with S-Max.

Conclusions: Cadence-controlled selective stimulation improves cycling endurance and increases efficiency over conventional stimulation by incorporating fiber groups only as needed to maintain a desired exercise intensity. Closed-loop carousel stimulation also successfully reduces power fluctuations relative to previous open-loop efforts, which will enable neuroprosthesis recipients to better take advantage of duty cycle reducing patterns.

Keywords: Cycling; Exercise; Feedback control; Neural stimulation; Paralysis; Spinal cord injury.

Fig. 1

Neural stimulation-driven cycling exercise setup

Fig. 2

Standard Controlled (S-Cont) stimulation schematic. Instantaneous cadence is calculated from moving-average filtered time derivative of the crank angle and compared against a target cadence. An error e(t) between target and instantaneous cadence drives one PI controller per leg to adjust the PW delivered through all knee extensor-activating contacts during the respective left and right active periods of quadriceps activity during the stroke cycle

Fig. 3

Carousel Controlled stimulation schematic for the left quadriceps (green box). Carousel logic (blue box) detects the passing of an angle (θ = 5) outside of the L. Quad active region and switches the stimulating contact and thus fiber group active in the next contraction. Instantaneous cadence and resulting errors are calculated as in the Standard Controller. Each independent contact is driven by its own independent PI controller, enabling different PW outputs to be delivered through the different contacts when active. When not active, contacts receive a PW of zero and thus do not contribute to the pedal stroke. This logic is repeated for the right quadriceps, using a different crank angle (θ = 180) as the contact switching signal

Fig. 4

Difference in W and P_end between controlled conditions and S-Max stimulation trials. Positive differences indicate improved outcomes compared with conventional, open-loop cycling. Percent improvement is given for differences with statistical significance (*p < 0.05). Note participant P02 only completed 2 trials of the S-Cont condition due to time constraints. All other participants completed at least three trials of cadence-controlled conditions and a corresponding number of S-Max trials

Fig. 5

Power fluctuation indices (PFI) for conventional and cadence-controlled stimulation conditions. Lower PFI indicates a smoother, more stable ride

Fig. 6

Example (a) S-Cont and (b) C-Cont PW output throughout a cycling trial for the same participant (P04). Colors indicate PWs delivered through independently-controlled electrode contacts. PW was adjusted as needed within each active contraction to maintain a target cadence, up to a hardware limited 255 µs maximum

Fig. 7

P05 muscle oxygenation (SmO₂) over time in the left vastus lateralis (top), left vastus medialis (middle), and left rectus femoris (bottom) throughout cycling trials with S-Max (red) and S-Cont (purple) stimulation. Shaded regions represent standard deviations. Cadence control enables higher SmO₂ values to be maintained in both the vastus lateralis and rectus femoris fiber groups during the first minute of exercise

Fig. 8

P05 average heart rate during S-Max and S-Cont stimulation-induced cycling bouts