There are many secondary functions associated with designing an electrocardiogram (ECG) signal chain, but one of the most important might be the most underrated: making sure that the electrodes are properly connected. Sounds simple enough – just stick the electrode on the patient and observe whether it stays on, right?
Physicians and cardiologists typically rely on an automated detection system to indicate when an electrode is no longer properly connected. This can be as simple as an on/off declaration for the most basic patient monitoring. However, accurately measuring how well the electrode is connected requires a little more finesse.
Contact impedance is typically higher for dry electrodes, but even wet electrodes can dry out over time. As electrodes become drier, the skin-to-electrode contact impedance increases, reducing signal amplitude and overall measurement resolution. Trained cardiologists may prefer to determine the quality of the electrode contact themselves before it becomes completely disconnected. So how can you accurately measure this impedance to give reliable lead-off detection in ECG systems?
How to implement basic lead-off detection
Lead-off detection requires applying a small current, either DC or AC, through the signal chain that passes through the ECG cables, electrodes and patient. The current produces a voltage proportional to the total impedance in the signal path. The current can be applied either by biasing each input with a high-impedance connection to a supply voltage or by connecting a controlled current source. The TI ADS1298 and related family of analog front-ends, for biopotential measurements, implements both current-application options, as shown in Figure 1.
Resistor biasing and current source excitation options for lead-off detection
The simplest lead-off detection method uses a comparator to monitor the DC voltage on each electrode input. Normally, electrode input signals are biased to a mid-supply common-mode voltage by a dedicated right-leg drive (RLD) electrode. If an input electrode becomes disconnected, the biasing resistor or current source will pull the DC voltage to one of the supply rails. The comparator, shown in Figure 2, then provides a binary indication of when the voltage exceeds a predetermined threshold. This method gives an immediate indication of a disconnected electrode. For further analysis, the voltage can be passed through a programmable gain amplifier (PGA) and converted by an analog-to-digital converter (ADC).
Monitor comparator outputs for on/off declaration; analyze ADC output for impedance measurement
How to measure the electrode impedance at DC
In order to determine the impedance of the signal path, the voltage produced by the lead-off current must be measured and converted into a digital code by an ADC. Translating these digital codes back into volts and dividing by the lead-off current yields the measured impedance. Any other known impedance, such as filter or protection components, must be subtracted from the result in order to isolate the electrode and cable impedance.
For DC current, all cable and electrode capacitance appear as an open circuit, so only the resistance of the signal path is measured. The voltage from the applied lead-off current will add to the DC offset voltage of the signal chain. This is acceptable for systems with sufficient dynamic range as the DC voltage can be removed from the ECG in post processing.
How to measure the electrode impedance at AC
At AC, electrode and cable capacitance is included in the impedance measurement. The AC current produces a voltage waveform whose peak amplitude is proportional to the magnitude of the impedance at the set lead-off detection frequency. The frequency of the AC lead-off current can vary depending on the application’s signal bandwidth of interest (typically 0.05 Hz-150 Hz for ECG) so as to enable simultaneous ECG and electrode impedance measurements.
Systems that must analyze both the magnitude and phase of the returned lead-off voltage signal can use in-phase/quadrature (I/Q) demodulation, which helps distinguish the total capacitance in the signal path from the resistance.
Calculating the ECG signal-path impedance
To calculate the magnitude of the impedance with pull-up or pull-down resistors, refer to Figure 3 and Equation 1:
DC impedance measurement with biasing resistors
To calculate the magnitude of the impedance with lead-off current sources, refer to Figure 4 and Equation 2:
DC impedance measurement with lead-off current sources
Here are some tips for improving electrode impedance measurement accuracy:
- Calibrate the current applied to each input. Semiconductor manufacturers will typically give a conservative tolerance for their lead-off current sources or biasing resistors. You can improve the accuracy of impedance measurements by using a precision resistance or multimeter to directly measure the current.
- Use the largest allowable magnitude of lead-off current in order to dominate other sources of error, especially unaccounted leakage paths.
- Understand the path that the current is taking to ensure that you are calculating the correct impedance. For instance, if current is only applied to the positive input electrode and not the negative input electrode, then the return path for the current may pass through the right-leg drive amplifier, as shown in Figure 5. The measured impedance will not include the negative channel input.
Current path may return through RLD electrode
- (DC only) Calibrate the DC offset voltage of the signal chain to remove other DC errors from the measurement. These errors include the offset voltages from active components as well as other components with significant leakage current. To perform calibration, short the inputs of each ECG channel on the printed circuit board and record the average result at the ADC output to subtract from the impedance measurements.
- (AC only) Remember to account for any filter attenuation at the lead-off current frequency. You should design your analog antialiasing filters with much higher cutoff frequencies than the signal bandwidth of interest, but be aware that digital filters in delta-sigma ADCs can introduce a significant gain error term depending on their response. You must divide the measured voltage amplitude by the digital filter gain at that frequency before calculating the impedance.
- (AC only) Apply a band-pass filter around the lead-off current frequency to remove the noise that may affect the measured signal amplitude.
We hope we’ve made clear the differences between monitoring electrode connection status and measuring electrode impedance to determine connection quality. You can use DC current for the most basic on/off declarations and to measure the electrode resistance, and use AC current to measure the total electrode impedance, including capacitive elements. Using the tips and tricks outlined here, it’s often possible to get more accurate measurements than specified in a semiconductor’s device data sheet.
About the authors
Alex Smith, applications engineer for Precision ADCs, Texas Instruments
Alex Smith is an applications engineer for Precision ADCs at Texas Instruments, where he supports the Biopotential, Isolated Amplifiers and Modulators and Sensor Measurement devices. Alex earned his Bachelor of Science in electrical engineering from the University of Arkansas in Fayetteville, AR.
Ryan Andrews, applications engineer, Precision ADCs, Texas Instruments
Ryan Andrews is an applications engineer for Precision ADCs at Texas Instruments where he supports the Biopotential, Motor Control and other consumer sectors. Ryan earned his Bachelor of Science in biomedical engineering and his Bachelor of Arts in Spanish from the University of Rhode Island in Kingston, RI.