Integrating Cardiac Cycle Phases into Event Related Potential Analysis: A Protocol for Two-Dimensional and Virtual Reality Environments

Summary

This protocol describes an approach to investigate cardiac phase-dependent modulation of visual event-related potentials utilizing simultaneous electroencephalography and electrocardiography recordings. The protocol details stimulus synchronization, cardiac phase segmentation, and event-related potentials analysis during systole and diastole, and is applicable in both conventional two-dimensional and immersive virtual reality environments.

Abstract

The heartbeat is a fundamental physiological rhythm that not only sustains life but also modulates perceptual and neural processes, thereby influencing the reproducibility of event-related potentials (ERPs). While virtual reality (VR) offers highly immersive environments compared to conventional two-dimensional (2D) screens, it also introduces unique challenges for electrophysiological recording, including motion artifacts, presentation latency, and potentially stronger cardiogenic modulation. These factors can diminish the efficiency of standard electroencephalography (EEG) preprocessing, leading to substantial data loss. This study presents a detailed protocol that combines simultaneous EEG and electrocardiography (ECG) to investigate heart-brain interactions under both VR and 2D conditions. The approach involves precise alignment of stimuli with cardiac signals, correction for VR-induced latency, and extended artifact management. Segmentation of EEG epochs based on whether stimuli occurred during cardiac systole or diastole revealed a robust suppression of visual ERP amplitude during systole. Critically, this suppressive effect was nearly identical in both VR and 2D environments, demonstrating the protocol's efficacy and confirming that this fundamental heart-brain interaction is resilient to changes in immersion level. This protocol, thus, offers a standardized framework for investigating cardiac influence on cognition, adaptable to diverse visual presentation methods.

Introduction

The rhythmic activity of the heart is increasingly recognized not only as a vital homeostatic process but as a fundamental modulator of brain function and cognition¹. A growing body of evidence demonstrates that cardiac signals exert a continuous influence on a wide spectrum of neurocognitive processes, including perception, attention, emotional regulation, and consciousness^2,3. This heart-brain axis is mediated by afferent interoceptive signals from cardiovascular baroreceptors and mechanoreceptors, which project via the brainstem to cortical areas where they integrate with exteroceptive sensory information^4,5. Consequently, the brain does not process external stimuli passively, but does so within the context of the body's internal state, a key indicator of which is the phase of the cardiac cycle.

The central methodological challenge is to reveal how cardiogenic neural activity influences the processing of external stimuli, as measured by visual ERPs. The most prominent electrophysiological manifestation of this interplay is the heartbeat-evoked potential (HEP), a specific pattern of EEG activity that is time-locked to the cardiac contraction⁶. The HEP is considered a neural marker of interoceptive attention, reflecting the brain's predictive coding of the impending heartbeat and its consequences for sensory processing^7,8. According to contemporary models, the brain utilizes these cardiac signals to create a unified representation of the body's internal state within its external context^9,10. Empirical evidence indicates that the processing of external stimuli, including their perceptual salience and underlying neural correlates, varies significantly depending on whether they are presented during cardiac systole or diastole^11,12,13. Stimuli presented during systole (the contraction phase) are consistently associated with suppressed neural processing, evidenced by reduced ERP amplitudes and altered behavioral responses, compared to those presented during diastole (the relaxation phase)^14,15,16. This effect is explained by the "somatosensory gating" hypothesis, which posits that the brain suppresses exteroceptive signal processing to temporarily prioritize cardiac interoceptive signals during systole^15,17.

Despite growing interest in the heart-brain axis, standard ERP analysis protocols focus exclusively on external stimuli, segmenting epochs relative to stimulus onset and thereby ignoring the profound trial-by-trial variability introduced by the ongoing cardiac cycle. To accurately capture cardiac-cycle modulation of visual ERPs, a fundamentally different approach is required. This necessitates the precise alignment of stimulus presentation with specific cardiac phases, verified by simultaneous ECG recording¹⁸. The current literature lacks standardized, detailed, and easily reproducible protocols that comprehensively describe all processing stages, from raw data acquisition to statistical analysis, accounting for cardiac phases.

An additional methodological challenge arises with the rapid adoption of immersive technologies like VR in neuroscience. VR environments provide highly immersive and engaging contexts, enabling the creation of controlled yet realistic experimental settings^19,20 . However, the simultaneous acquisition of high-quality EEG and ECG data within a VR environment is associated with several technical difficulties. These include potential motion artifacts, electromagnetic interference from the headset, and the critical need for precise temporal synchronization between stimulus presentation (often with inherent latency), data acquisition, and motion tracking systems²¹. These technical challenges have contributed to a scarcity of robust analysis protocols specifically designed for direct comparison of cardio-dependent neural modulation between classical 2D and immersive VR paradigms. This lack of a common framework ultimately hinders research into how immersion and presence modulate fundamental brain-body interaction mechanisms.

To address these challenges, this work presents a comprehensive, step-by-step methodological protocol for investigating cardiac-cycle modulation of visual ERPs by integrating simultaneous EEG and ECG recordings. The cornerstone of the approach is the segmentation of neural data relative to the cardiac R-peak, allowing for the classification of stimuli based on their occurrence during systole or diastole. Crucially, a data-driven analysis of the entire post-stimulus epoch is employed to identify specific time windows where cardiac phase exerts a significant modulatory influence on the neural response. The primary contributions of this work are threefold: (1) a complete, visualized protocol designed for straightforward replication is provided, (2) the protocol is adapted and validated to compare data acquired using a standard 2D monitor and an immersive VR headset, and (3) the utility of the protocol is demonstrated by revealing how facial stimuli of emotional valence differentially modulate cardiac influence on neural processing depending on the immersiveness of the environment.

Protocol

The study was approved by the Bioethics Committee of the Samara State Medical University (protocol code 195, dated 10 October 2018). All participants provided informed voluntary consent prior to the study. The final sample consisted of 27 healthy male volunteers aged 19-21 years. During the testing sessions, each participant completed a series of standardized assessments administered individually. The reagents and the equipment used are listed in the Table of Materials.

1. Materials and equipment

NOTE: Use functionally equivalent hardware and software when appropriate. Refer to Figure 1 for an overview of the experimental setup, including system layout and connections. Prepare the following equipment and software prior to the experiment. General requirements are described below, and specific configurations are listed in the Table of Materials.

1. EEG/ECG acquisition system
   1. Use an amplifier system that enables simultaneous recording of EEG and ECG.
   2. Use EEG electrodes that can be reassigned from the scalp for ECG recording.
2. ECG input options (EEG system supports bipolar channels)
   1. Use dedicated bipolar input channels for ECG recording, when available.
   2. Reassign two non-critical EEG electrodes to form a bipolar ECG pair if dedicated bipolar channels are unavailable.
3. ECG input options (EEG system supports monopolar channels only)
   1. Use a differential adapter to route ECG signals into an auxiliary (AUX) or differential input.
   2. Use passive electrodes and disposable adhesive electrodes compatible with the selected input configuration.
4. Stimulus presentation
   1. Use a computer with a high-refresh-rate display (≥120 Hz) for 2D presentation.
   2. Use a head-mounted display with a secondary monitor for signal mirroring for VR presentation.
5. Synchronization and sensors
   1. Use a photodiode or photosensor with mounting accessories and connect it to the EEG acquisition system.
6. Software
   1. Use a VR runtime environment and stimulus control software capable of precise timing, sequencing, and trigger output.
   2. Use EEG/ECG analysis software that supports event handling, preprocessing, and ERP analysis; use optional custom scripts if required.
7. Laboratory environment and accessories
   1. Use a shielded and properly isolated room for EEG recording.
   2. Provide a comfortable chair, cable management solutions, skin preparation supplies, conductive gel, and hygiene materials for head-mounted displays.

2. System setup and latency detection

1. Install and configure the stimulus presentation software.
   1. VR condition
      1. Install the vendor-specific VR runtime required for the head-mounted display.
      2. Deploy the experimental stimulus application to the VR headset according to the runtime documentation.
      3. Perform the initial VR system setup by calibrating the tracking system (if applicable) and defining the play area boundaries according to the manufacturer’s instructions to ensure stable tracking and minimize motion-related artifacts.
   2. 2D condition
      1. Optimize the stimulus presentation workstation by disabling unnecessary background applications and processes to minimize interruptions and timing delays during data acquisition.
2. Create the recording montage in the EEG acquisition software.
   1. Create a new workspace or recording configuration and name it according to the study identifier.
   2. Assign EEG channels and reserve AUX input channels according to the amplifier hardware documentation.
   3. Add two additional channels for the photodiode and ECG signals. Configure channel names, types, and units as appropriate, and include these channels in the same recording file as the EEG. Label the ECG channel consistently (e.g., “ECG”).
3. Measure stimulus presentation latency using a photodiode.
   1. Connect the photodiode to an available AUX input of the EEG amplifier.
   2. VR condition: Mount the photodiode on the headset lens at a location corresponding to a stimulus-locked luminance patch. Present at least 100 stimulus events and record the photodiode signal to measure display latency and trial-to-trial jitter relative to the intended trigger. After completing the VR measurements, relocate the photodiode to the 2D display.
   3. 2D condition: Place the photodiode over a dedicated stimulus-locked pixel area on the monitor. Present at least 100 stimulus events and record the photodiode signal to measure display latency and jitter relative to the intended trigger.
   4. Optional: If two photodiodes are available, connect each sensor to a separate AUX channel and present synchronized flashes to measure 2D and VR latencies simultaneously.
   5. Remove the photodiode from the 2D display after completing latency measurements to avoid obstructing the participant’s view.
   6. Compute the mean display latency and the standard deviation (jitter) for each presentation environment.
   7. Document latency and jitter values separately for the 2D and VR conditions.
   8. Verify that latency jitter remains stable across the session. If hardware or software settings are modified, repeat the latency measurements.
4. Configure the ECG recording channel.
   1. Determine the ECG input configuration based on the available amplifier capabilities (bipolar channels supported vs. monopolar channels only; see steps 1.2 and 1.3).
   2. If bipolar ECG recording is supported, enable the dedicated bipolar ECG input in the recording software and reserve the channel label “ECG.”
   3. If only monopolar channels are available, connect a differential adapter to an AUX or differential input and map this input as the ECG channel in the recording software. Reserve the channel label “ECG.”
   4. Do not place ECG electrodes at this stage. Perform electrode placement and impedance checks during participant preparation.

3. Participant preparation

1. EEG cap placement
   1. Fit the EEG cap according to the international 10–10 or 10–20 system by aligning the Cz electrode and adjusting the cap size to the participant’s head circumference²².
   2. Part the hair at each electrode site using a cotton swab or a blunt-tipped cannula (approximately 2–3 mm diameter). Apply a mild abrasive or skin preparation solution to reduce impedance.
   3. Place each electrode, apply conductive gel, and ensure firm contact with the scalp by gently rotating the electrode holder. Avoid creating conductive gel bridges between neighboring electrodes.
   4. Measure electrode impedances and reduce them below 8 kΩ (or the laboratory standard) by repeating hair parting, applying additional skin preparation, or adding conductive gel as needed.
   5. Secure EEG cables to minimize movement and verify that the participant feels comfortable before proceeding.
2. ECG electrode placement
   1. Clean the skin on the inner (palm-facing) side of both wrists using alcohol wipes and allow the skin to dry completely.
   2. Place disposable adhesive electrodes on the wrists to record Standard Limb Lead I, positioning the negative electrode on the right wrist and the positive electrode on the left wrist. Ensure good electrode contact and avoid placement over bony prominences.
   3. Secure ECG cables using medical tape, leaving small slack loops to reduce motion-related artifacts.
   4. Verify the ECG signal at rest. Ensure that R-peaks are clearly visible and upright. If R-peaks appear inverted, swap the electrode polarity. Confirm that the ECG signal remains stable over time.
   5. If using a single amplifier for EEG and ECG recording, connect the EEG and ECG ground electrodes as required by the system. When multiple ground contacts are available, place separate ground electrodes for EEG and ECG to improve signal stability, ensuring that all ground electrodes are internally connected within the amplifier.

4. VR headset fitting (VR condition only)

1. Do not remove the EEG cap or disconnect any EEG or ECG electrodes prior to headset fitting.
2. Loosen the top strap and the rear tightening mechanism of the VR headset to accommodate placement over the EEG cap.
3. Instruct the participant to position the front section of the VR headset on their face. Lower the rear section over the EEG cap and gradually tighten the straps until the headset is securely fitted.
   1. Ensure that the headset does not exert excessive pressure on EEG electrodes. Adjust the fit to prevent headset slippage while maintaining participant comfort during prolonged use. Inspect the EEG signal for line noise or pressure-induced artifacts and readjust the headset if necessary²³.
4. Mirror the VR display output to a secondary monitor to allow real-time supervision of stimulus presentation and participant behavior.
5. Recheck EEG electrode impedances after headset fitting. Reapply conductive gel to any channel exceeding 8 kΩ (or the laboratory standard). Repeat impedance checks periodically throughout the VR condition.

5. Recording

1. Stimulus presentation
   1. Present visual stimuli with precise and reproducible timing. Send and record event triggers for each stimulus onset within the EEG recording. Use identical stimulus timing parameters (e.g., stimulus duration, inter-stimulus interval) across the 2D and VR environments.
2. Participant monitoring
   1. Instruct the participant to minimize head movement and eye blinking during stimulus presentation. Continuously monitor the EEG and ECG signals during recording. Pause the experiment if necessary to correct electrode impedances or adjust cabling.
3. Recording blocks
   1. Record data in both presentation environments (2D and VR). Counterbalance the order of the environments across participants when applicable to control for order effects.

6. EEG preprocessing

1. Data import and channel verification
   1. Import the continuous EEG, ECG, and photodiode data into the EEG analysis software. Verify that all channel labels, types, and units are correctly assigned.
2. Signal filtering
   1. Apply a notch filter at the local power line frequency (50 Hz or 60 Hz). Include harmonic frequencies if required by the laboratory standard.
   2. Apply a band-pass filter to the EEG data according to the laboratory standard (e.g., 0.1–40 Hz).
3. Event marker correction and relabeling
   1. Correct stimulus event markers by subtracting the environment-specific display latency measured in step 2.3. Generate latency-corrected stimulus markers for subsequent analyses.
   2. Relabel stimulus markers according to the presentation environment (e.g., prefix event labels with “2D_” or “VR_”).
4. Artifact detection and correction
   1. Inspect visually the continuous data for artifacts. Identify and mark segments containing excessive noise or movement artifacts. Interpolate bad channels when appropriate, following standard laboratory procedures.
   2. Perform independent component analysis (ICA)²⁴ to identify ocular and muscle artifacts. Inspect component time courses, scalp topographies, and spectral characteristics. Remove artifact-related components based on objective criteria or expert judgment. When appropriate, extend ICA- or regression-based approaches to model cardiac-related artifacts.
   3. Compute ERPs time-locked to the R-peak using the same preprocessing parameters as for stimulus-locked ERPs. Inspect the resulting waveforms and scalp topographies to assess potential cardiac field artifact (CFA) contamination²⁵. Identify prominent QRS-like complexes or other ECG-related deflections in the EEG signals (Figure 2), that indicate residual cardiac activity volume-conducted to the scalp.
      1. If substantial cardiac-related activity is observed, apply additional correction procedures, general linear model-based regression using ECG as a continuous predictor²⁶, extended ICA-based methods, or machine learning–based approaches²⁷.
   4. Account for overlap between stimulus-locked ERPs and HEP. Use methods that explicitly model overlapping neural responses, such as regression-based or deconvolution approaches²⁶. When appropriate, apply surrogate-based procedures or control analyses to assess the robustness of cardiac-related neural effects^28,29.

7. Cardiac phase segmentation (offline)

1. Detect R-peaks
   1. Detect R-peaks automatically in the ECG signal using the analysis software. Manually review all detected events and correct missed or falsely detected R-peaks to ensure accurate heartbeat timing.
2. Define cardiac phases
   1. Define systole and diastole while accounting for inter-individual differences in heart rate and beat-to-beat variability. Prefer cycle-specific definitions over fixed temporal windows when possible.
      NOTE: Use one of the following approaches: (1) Define systole and diastole as proportions of each individual R-R interval, consistent with the non-linear relationship between heart rate and cardiac phase durations; (2) When ECG signal quality permits, define the end of systole using the end of the T-wave as a physiologically grounded marker; (3) If reliable T-wave detection is not feasible, define systole using a fixed temporal window (e.g., the first 300 ms following the R-peak). Explicitly acknowledge this approximation and account for individual heart rate differences in subsequent analyses.
3. Assign cardiac phase labels to stimuli
   1. Use latency-corrected stimulus markers to assign each stimulus to the cardiac phase present at stimulus onset (systole or diastole). Assign phase labels separately for each presentation environment (2D or VR) and rename events accordingly (e.g., 2D_systole, VR_diastole).
4. Exclude ambiguous trials
   1. Identify and remove trials in which the planned ERP analysis window would span both cardiac phases based on the participant’s instantaneous heart rate.
5. Example software implementation
   NOTE: The following steps illustrate one possible implementation using an EEG analysis software package that supports ECG event detection and event relabeling.
   1. Automated R-peak detection: Open the event detection menu and select the heartbeat detection function. Select the ECG channel with the highest signal quality, define a label for the detected events (e.g., “R-peak”), and run the detection algorithm to generate R-peak event markers.
   2. Manual review of R-peak detection: Inspect the ECG signal visually with overlaid R-peak markers. Add missing markers and remove incorrect detections using the event editing tools provided by the software.
   3. Cardiac phase-based event relabeling: Use the event combination or selection function to classify stimuli according to cardiac phase at onset.
      ​NOTE: For systolic trials, select stimuli occurring within the predefined systolic interval following each R-peak. For diastolic trials, select stimuli occurring within the predefined diastolic interval preceding the next R-peak. Create new event categories for systole and diastole. Perform this procedure separately for each cardiac phase and presentation environment.

8. Epoching and averaging

1. Epoching and baseline correction
   1. Assign cardiac phase labels exclusively based on the cardiac phase present at stimulus onset (i.e., systole or diastole). Define the ERP epoch relative to stimulus onset. Allow the epoch to extend into subsequent cardiac phases without affecting phase classification.
   2. Use an epoch length of approximately 400 ms. If a longer window (e.g., 600 ms) is required, exclude trials in which the analysis window extends substantially into the subsequent cardiac phase. Apply baseline correction using a short pre-stimulus interval (e.g., -50 to 0 ms, or -50 to -2 ms).
2. Averaging
   1. Compute ERPs separately for the four experimental conditions: 2D_systole, 2D_diastole, VR_systole, and VR_diastole. Report the number of retained trials for each condition.
3. Electrode selection considerations
   1. Select electrodes for analysis based on the research question and the nature of the ERP components of interest. Use occipital electrodes (O1, Oz, O2) for visual paradigms as they capture canonical visual ERPs with high signal-to-noise ratio.

9. Quality control and troubleshooting

1. Trigger-stimulus latency jitter
   1. Assess trial-to-trial latency jitter between trigger timestamps and photodiode-measured stimulus onset. Consider jitter exceeding approximately 10 ms across trials as problematic, particularly for early ERP components. If excessive jitter is observed, re-evaluate stimulus timing, hardware synchronization, and apply run-specific latency corrections as needed.
2. ECG saturation or noise
   1. Check ECG electrode contact and impedance. Swap electrode polarity if R-peaks are inverted. Relocate ECG electrodes to the forearms if wrist recordings remain noisy.
3. VR headset pressure artifacts
   1. Refit headset straps to reduce pressure on EEG electrodes. Add padding if necessary and reapply conductive gel to affected electrodes.
4. Unequal trial counts across cardiac phases
   1. Adjust the ERP epoch window or stimulus timing to ensure balanced sampling across systolic and diastolic phases.
5. Ambiguous-phase trials
   1. Identify trials in which stimulus onset occurs near a cardiac phase boundary (e.g., around the R-peak or near the transition between systole and diastole) and where the planned analysis window may span both phases. Flag these trials as ambiguous. Inspect the ECG signal and the instantaneous R-R interval and exclude ambiguous trials from further analysis.

Results

The present study employed an analysis protocol to investigate ERP modulation by afferent interoceptive signals during the cardiac cycle phase (systole vs. diastole) across two experimental environments: a standard 2D monitor and an immersive VR setting. EEG epochs were extracted and analyzed relative to the R-peak of the concurrently recorded ECG.

The nonparametric cluster-based permutation test applied to visual ERPs recorded at occipital electrodes (O1, Oz, O2) across the entire epoch revea...

Discussion

Critical steps within the protocol
Several steps of the proposed protocol are critical for obtaining valid cardiac phase-dependent ERP effects. First, precise synchronization between stimulus presentation and EEG recording is essential, and environment-specific display latency must be quantified using a photodiode and corrected at the event-marker level, as uncorrected delays or excessive jitter can lead to temporal smearing of ERP components. Second, accurate detection of ECG R-peaks and reliable ...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
actiCAP slim / snap	Brain Products GmbH	BP-185-2200	Active electrode cap with 64 channels.
actiCHamp Plus Versatile all-in-one lab amplifier	Brain Products GmbH	S/N ACBM17120955	Primary EEG data acquisition system.
Ag/AgCl electodes disposable 26 mm	Nika-Medical Ltd	N/A	Disposable electrodes for three-lead ECG
BIP2AUX adapter	Brain Products GmbH	N/A	Connects ECG output to EEG aux input.
Desktop ROG Srtix	ASUSTeK Computer Inc.	G835LR-SA117	High-performance workstation. Used for simultaneous operation of the VR system (HTC Vive), stimulus presentation software, and EEG recording platform (BrainVision Recorder).
HTC Vive  Pro Eye	HTC Corporation	S/N FA1142100443	VR system for immersive stimulus presentation.
Photo Sensor	Brain Products GmbH	N/A	For optical detection of stimulus onset.
Samsung 27" ViewFinity S8	Samsung Electronics Co., Ltd	S/N 0UYMHNBXB00044D	Primary display for the 2D condition.
Samsung 27" ViewFinity S8	Samsung Electronics Co., Ltd	S/N 0UYMHNBXB00030E	Experimenter's display. Used for running EEG recording software and monitoring data during experiments.
Shielded-room Neiroiconica "Expert"	Neiroiconica Assistive Ltd	S/N K20210119011	Electromagnetically shielded chamber.
Snap Lead Cable	Neurosoft ltd	N/A	Connects ECG electrodes to the adapter
TriggerBox Plus	Brain Products GmbH	S/N TB-A100009-0624	Synchronization interface for TTL triggers.