Gradient Biotech
U

Menu

Research Areas

Computational BiologySingle-cell, bulk RNA-seq, spatial, and biomarkers
CardiologyCardiovascular signals, cohorts, and interpretationImmunologyImmune cell states, signaling networks, and repertoire dynamicsNeurologyBrain connectivity, functional imaging, and electrophysiologyOncologyTumor microenvironment, immune response, mutations, and survivalPathologyWhole-slide imaging, tissue quantification, and spatial pathology

Application

Documentation

Documentation

Back to studies

Multimodal autonomic profiling

Combine ECG, blood pressure, and respiration recordings for a single subject — compute cross-modal coupling metrics and generate mechanistic AI interpretation.

Research question

How do heart rate variability, blood pressure dynamics, and respiration interact for this participant? What do baroreflex sensitivity and respiratory sinus arrhythmia reveal about autonomic regulation?

Who this is for

  • Cardiovascular physiology labs studying autonomic control mechanisms
  • Translational research programs with synchronized ECG, BP, and respiration recordings
  • Hospital research groups exploring baroreflex and RSA in hypertension or heart failure cohorts

Data requirements

DataRequiredPurpose
ECG or RR recordingYesHRV metrics for coupling pipelines
Blood pressure CSVYes (for baroreflex)SBP/DBP series, variability, dipping
Respiration channelYes (for RSA)Breath timing and rate for RSA computation

All datasets must be linked to the same subject in the subjects table.

Workflow

Import subject → Upload RR/ECG + BP + respiration → HRV + analyze BP → Baroreflex / RSA → Mechanistic interpret

Step 1 — Link multimodal datasets

Import the subject and upload modality-specific datasets:

  • rr_p001.csv or ECG waveform → HRV pipeline
  • bp_p001.csv → blood pressure analysis
  • Respiration recording → respiration analysis (when available)

Assign all datasets to subject P001 from the subjects page.

Step 2 — Run single-modality pipelines

From the Subject hub (/experiments/{id}/subjects/{subjectId}):

  1. Run Compute HRV on the RR/ECG dataset
  2. Run Analyze BP on the blood pressure dataset — review SBP/DBP/MAP means, variability, and dipping classification
  3. Run Analyze respiration on the respiration dataset (for RSA)

Step 3 — Cross-modal coupling

Launch coupling pipelines from the subject hub when prerequisite runs complete:

PipelineInputsOutputs
Compute baroreflexHRV/RR run + BP runBaroreflex sensitivity (sequence method)
Compute RSAECG preprocess run + respiration runRespiratory sinus arrhythmia metrics

Step 4 — PPG path (optional)

When PPG rather than ECG is available:

  1. Preprocess PPG → pulse peaks and IBI intervals
  2. Compute PRV → pulse rate variability metrics

PRV can substitute for HRV in coupling analyses when appropriate.

Step 5 — Mechanistic interpret

Use Interpret (mechanistic) on the subject hub when HRV, BP, and coupling metrics exist. Mechanistic mode connects cross-domain findings into autonomic hypotheses citing computed values.

Expected outputs

  • HRV and BP variability metrics per subject
  • Dipping classification and MAP statistics
  • Baroreflex sensitivity estimate
  • RSA coupling metrics
  • Multimodal timeline view across linked recordings
  • Mechanistic AI narrative connecting HRV, BP, and respiration findings

Example insight

Patients with elevated LF/HF ratios and reduced RMSSD demonstrated significantly higher arrhythmia burden during nighttime telemetry, with attenuated baroreflex sensitivity suggesting impaired autonomic buffering.

Related guides