GalaxyPPG: Consumer-Grade Wearable PPG and ECG Dataset for Cardiovascular Monitoring (Scientific Data 2025)

Overview

The GalaxyPPG Dataset published in Scientific Data (2025) addresses a critical gap in wearable sensor research by providing data from consumer-grade devices alongside research-grade sensors, enabling realistic performance evaluation under motion and stress.

Participant and Protocol Details

• 24 participants with diverse demographics recruited for semi-naturalistic data collection.
• Each participant performed a standardized series of activities designed to cover resting states, physical exertion, cognitive stress, and motion artifacts.
• Activities included sitting quietly, standing, walking on a treadmill at varying speeds, cycling, performing mental arithmetic tasks, and engaging in free movement.
• Session duration typically 60-90 minutes per participant, capturing transitions between activity states.

Sensor Setup and Synchronization

• Samsung Galaxy Watch 5: Consumer smartwatch worn on the wrist, continuously recording PPG signals using built-in optical heart rate sensor (green LED photodiode array).
• Empatica E4: Research-grade wrist-worn device measuring PPG, skin temperature, electrodermal activity (EDA), and 3-axis accelerometer, serving as a quality benchmark.
• Polar H10: Chest-strap ECG monitor providing gold-standard heart rate and R-R interval measurements for ground truth validation.
• Hardware-level or software-level time synchronization ensuring alignment of all three data streams for multi-modal analysis.

Measured Signals

• PPG (Galaxy Watch & Empatica E4): Raw photoplethysmography waveforms sampled at device-specific rates (typically 25-64 Hz), capturing blood volume pulse.
• ECG (Polar H10): Single-lead chest ECG at 130 Hz or higher, enabling precise R-peak detection and heart rate variability (HRV) analysis.
• Accelerometer data: 3-axis motion measurements from wrist devices for motion artifact characterization and activity classification.
• Derived metrics: Heart rate, inter-beat intervals (IBI), HRV parameters (SDNN, RMSSD, pNN50), and signal quality indices.

Dataset Value Proposition

• First large-scale dataset comparing consumer-grade smartwatch PPG (Galaxy Watch) with research-grade E4 and ECG ground truth under identical conditions.
• Enables assessment of Galaxy Watch sensing performance across activity types, motion intensities, and stress levels.
• Supports development of robust PPG processing algorithms that account for motion artifacts, skin tone variations, and device placement differences.
• Facilitates validation of consumer wearable accuracy for clinical and wellness applications.

Use Cases

• Algorithm development: Training and validating PPG-based heart rate, HRV, and arrhythmia detection algorithms for consumer devices.
• Device benchmarking: Quantifying accuracy, precision, and reliability of consumer smartwatches vs. research-grade sensors.
• Motion artifact mitigation: Developing signal processing techniques to improve PPG quality during physical activity.
• Clinical validation: Assessing feasibility of consumer wearables for remote patient monitoring and cardiovascular disease screening.
• Personalized health: Exploring individual differences in PPG signal quality and deriving personalized calibration models.