Impedance-Derived Heart Rate and Heart Rate Variability from tDCS Output Voltage: Sensorless Physiological Monitoring During Electrical Neuromodulation

Abstract

BackgroundTranscranial direct current stimulation (tDCS) devices adjust output voltage to maintain the target current despite varying impedance. Pulsatile blood flow produces beat-synchronous changes in tissue impedance. ObjectiveTo determine whether impedance-derived heart rate (IHR), heart rate variability (IHRV), and respiration (IDR) can be estimated from tDCS output voltage without additional physiological sensors. MethodsA custom analog front-end acquires the tDCS output voltage across its full dynamic DC range and superimposed AC fluctuations with high precision. Beats detected from the AC-coupled signal yielded normal-to-normal intervals for HR, HRV, and interval-derived respiration. Accuracy was quantified as mean absolute error (MAE) in 10 healthy laboratory participants against ECG and respiration-monitor references, and against chest-strap RR intervals in 19 at-home sessions from 10 participants with mild-to-moderate depression. ResultsLaboratory MAEs versus ECG were 0.57 bpm for HR, 9.40 ms for SDNN, and 18.90 ms for RMSSD (r = 0.995, 0.859, and 0.778; N = 10); respiratory-rate MAE was 1.36 breaths/min (r = 0.852; N = 6). Across 1-5 mA of tDCS, the cardiac voltage {Delta}Vcardiac(t) amplitude scaled linearly with current (slope, 0.073 mV/mA; p < 0.001). The pulsatile impedance {Delta}Zcardiac(t) = ({Delta}Vcardiac(t)/Iapplied) amplitude averaged 0.080 {+/-} 0.029 {Omega} (mean {+/-} SD) across 50 participant-current observations, with no significant dependence on intensity (slope, -0.002 {Omega}/mA; p = 0.086). At-home MAEs were 1.43 bpm for HR, 8.86 ms for SDNN, and 24.92 ms for RMSSD (r = 0.995, 0.767, and 0.680; 19 sessions). ConclusionstDCS output voltage contains a recoverable cardiac-synchronous signal arising from pulsatile impedance, enabling HR, HRV, and respiratory monitoring without additional physiological sensors.