KN95 Mask Respirometer for COPD

By /

KN95 Mask Respirometer for COPD

ISEF Category: Biomedical Engineering

Ready to Turn This Idea Into a Real Project?

This guide was put together with the help of AI research tools to give you a solid starting point. But a competitive science fair project lives in the details: refining your research question, fine-tuning your variables, analyzing your data, and presenting your findings like a seasoned scientist.

For next steps tailored to your interests, skill level, and timeline, work one-on-one with a MehtA+ mentor. Learn more about MehtA+ Science & Engineering Research Mentorship →

Subcategory: Biomedical Devices  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

Hospitals send COPD patients home with a peak-flow meter from the 1980s. Most people forget to use it. A clip-on respirometer that lives on a KN95 mask captures every breath, all day, with no extra effort. A small ML model spots the slow drift that warns of an exacerbation before it lands the patient back in the ER.

What Is It?

A differential-pressure sensor measures the small pressure drop across the mask filter. With a calibration curve, that pressure converts to airflow.

Integrating airflow over each breath gives tidal volume. Tracking tidal volume, breath rate, and work-of-breathing over days produces personalized baselines.

A classification ML model trained on PhysioNet respiratory data flags departures from the personal baseline that match known exacerbation patterns. The output is a probability score, not a diagnosis.

Why This Is a Good Topic

Remote respiratory monitoring is a real clinical problem and the hardware fits inside a mask. You will learn pneumatic sensing, baseline modeling, and clinical-alert design.

Research Questions

  • How does mask seal change measured tidal volume?
  • What is the effect of breath rate on calibration accuracy?
  • Does the early-warning model beat a fixed-threshold baseline?
  • To what extent does activity level confound the signal?
  • Which feature carries the most exacerbation information?
  • How does mask filter degradation shift readings?
  • What is the effect of training-set size on per-subject calibration?

Basic Materials

  • MPXV7002 differential-pressure sensor.
  • ESP32 development board.
  • KN95 masks.
  • 3D-printed sensor clip.
  • Spirometer calibration syringe (low-cost).
  • LiPo battery.
  • Informed-consent template.

Advanced Materials

  • Clinical spirometer for ground truth.
  • Hospital mentor for COPD-cohort access.
  • IRB approval for patient recordings.
  • Higher-precision pressure transducer.

Software & Tools

  • PyTorch: Trains the exacerbation classifier.
  • PlatformIO: Builds the firmware.
  • Python (NumPy and SciPy): Processes breath-by-breath signals.
  • PhysioNet datasets: Provides training data.

Experiment Steps

  1. Calibrate the sensor with a known flow source.
  2. Lock the mask clip geometry and seal protocol.
  3. Decide features (tidal volume, breath rate, work-of-breathing) and label sources.
  4. Build a subject-wise data split.
  5. Train and validate the early-warning model.
  6. Report alert sensitivity, specificity, and per-subject calibration.

Common Pitfalls

  • Treating an uncalibrated pressure reading as flow.
  • Mixing mask seals between trials.
  • Ignoring activity-level confounds.
  • Training on too few subjects and claiming a general result.
  • Skipping a fixed-threshold baseline for comparison.

What Makes This Competitive

Calibrate the pressure sensor against a known flow source. A competitive project runs subject-wise data splits, reports calibration plots vs. spirometry where available, and benchmarks alert sensitivity and specificity against published COPD exacerbation studies.

Project Variations

  • Add a temperature sensor for breath-temperature trends.
  • Pair with pulse oximetry for an oxygen-saturation cross-check.
  • Build a similar clip for surgical masks and compare.

Learn More

  • PubMed: Search mask respirometer COPD review.
  • NIH PubMed Central: Open-access pulmonary monitoring papers.
  • PhysioNet Respiratory Database: Free with documentation.
  • American Thoracic Society: Open clinical resources.
  • MIT OpenCourseWare: Course 6.555 Biomedical Signal and Image Processing.
Shopping Cart