Chest Accelerometer Arterial Stiffness Project

ISEF Category: Physics and Astronomy

Subcategory: Biological Physics  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

Your heartbeat makes your body move in tiny ways, and a phone can sometimes pick up that motion. That means you can turn a chest accelerometer into a simple heart signal sensor. With careful signal processing, you can test whether pulse-related timing shifts with age or fitness. This is physics, physiology, and data analysis in one project.

What Is It?

Ballistocardiography, or BCG, is the study of tiny body movements caused by blood moving through the heart and large arteries. When your heart ejects blood, your body gets a small recoil. A sensitive accelerometer on the chest can record that motion. Think of it like feeling the pushback when a hose suddenly starts moving water, except the hose is your circulatory system.

The goal here is to turn that messy motion into a useful measurement. You can filter the signal, find repeating peaks, and compare timing features across people. One physics idea behind this is pulse wave velocity, which is the speed of the pressure wave traveling through the arteries. Stiffer arteries tend to send the wave faster. The Bramwell-Hill equation links that wave speed to arterial stiffness, so your project can test whether a simple motion signal tracks a real biomechanical property.

Why This Is a Good Topic

This is a strong science fair topic because you can measure something real from a home setup, then connect it to a well-known physics model. You get a clear independent variable, like age group, exercise level, or body position, and a measurable dependent variable, like pulse timing or estimated wave speed. The project also connects to a real health problem, since arterial stiffness relates to cardiovascular risk. You can learn signal processing, data cleaning, and model validation without needing a hospital lab.

Research Questions

• How does age group affect the timing between chest acceleration peaks and the cardiac cycle in ballistocardiography?
• What is the effect of body position on the repeatability of BCG peak timing measured with a chest accelerometer?
• Does recent light exercise change estimated pulse-wave-related timing compared with resting measurements?
• To what extent does fitness level correlate with BCG signal amplitude and signal-to-noise ratio?
• Which signal filter settings produce the most stable BCG peak detection across different participants?
• How does the pulse wave velocity estimated from BCG compare with values predicted by the Bramwell-Hill equation across participants?

Basic Materials

• Smartphone with Phyphox installed.
• Chest strap, snug athletic shirt, or adhesive phone mount to keep the sensor steady.
• Quiet chair or bed for seated and supine measurements.
• Stopwatch or timer for session tracking.
• Notebook or spreadsheet for recording participant age range, fitness proxy, and test conditions.
• Ruler or tape measure for estimating arterial path length if your protocol needs it.
• Consent form and parent or guardian approval for any minor participant.
• Alcohol wipes for cleaning any reusable contact surfaces.
• Digital scale or wall chart for optional body-size grouping.

Advanced Materials

• High-sensitivity accelerometer or wearable motion sensor with exported raw data.
• ECG device for timing reference and validation if your school or mentor allows it.
• Blood pressure cuff for additional vascular comparison.
• Pulse oximeter for basic heart-rate reference.
• Sampling interface or data logger for synchronized sensor recording.
• Calibration mass or vibration source for sensor response checks.
• Reference software for advanced spectral analysis and curve fitting.
• Institutional review documents if your site requires formal human-subject review.

Software & Tools

• Phyphox: Records accelerometer data from the phone and exports raw signals for analysis.
• Python: Cleans the signal, finds peaks, and compares participant groups.
• ImageJ: Helps inspect plots, screenshots, or signal traces when you need visual checks.
• Excel: Organizes participant metadata and summarizes basic statistics.
• MIT OpenCourseWare: Offers free signals and systems material if you need a refresher on filtering and Fourier ideas.

Experiment Steps

1. Define the physiological claim you want to test, such as whether age or fitness changes a BCG-derived timing feature.
2. Choose one sensor setup and keep it fixed, so differences come from people, not from changing hardware.
3. Plan how you will collect a clean resting baseline and how you will reject noisy trials.
4. Build a signal-processing workflow that finds consistent peaks, filters motion noise, and extracts one primary metric.
5. Decide how you will compare your metric with a physics model such as the Bramwell-Hill equation.
6. Set your statistics plan before you collect data, so you know how you will compare groups and report uncertainty.

Common Pitfalls

• Letting the phone shift on the chest, which adds motion artifacts that look like heart signals.
• Mixing up random body movement with true ballistocardiography peaks, which makes peak detection unreliable.
• Comparing participants with different measurement postures, which changes the signal more than age or fitness does.
• Using too many filters at once, which can erase the waveform shape you need for analysis.
• Treating noisy pulse-wave estimates as exact values, which hides uncertainty in the Bramwell-Hill comparison.

What Makes This Competitive

A strong version of this project does more than record a few heart traces. You would predefine a clean signal pipeline, compare multiple participant groups, and report uncertainty for every extracted feature. You could also test whether BCG-based estimates agree with another reference measure, or whether a specific signal feature predicts age or fitness better than simple heart rate. That kind of careful validation makes the work feel like real physics research, not just a demo.

Project Variations

• Compare seated, standing, and supine chest accelerometer signals to see how posture changes BCG peak timing.
• Test whether athletic students and non-athletic students differ in signal stability, pulse-wave estimates, or peak-to-peak variability.
• Add a second sensor, such as a smartwatch heart-rate trace, and compare its timing against the accelerometer-based signal.

Learn More

• PubMed: Search review articles on ballistocardiography, pulse wave velocity, and arterial stiffness to find background and validation studies.
• NIH PubMed Central: Read full-text physiology and biomedical engineering papers that are openly available.
• NASA NTRS: Search for signal processing and sensor analysis methods that can help with noisy motion data.
• MIT OpenCourseWare: Find free lectures on signals, filtering, and Fourier analysis for building your processing workflow.
• Biomedical Signal Processing and Control: Search recent journal articles on BCG feature extraction and wearable cardiovascular sensing.