Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Noncontact Respiratory Anomaly Detection Using Infrared Light-Wave Sensing

Md Zobaer Islam, Brenden Martin, Carly Gotcher, Tyler Martinez, John F. O'Hara, Sabit Ekin

TL;DR

This work demonstrates noncontact respiration monitoring with infrared light-wave sensing (LWS) and machine-learning-based anomaly detection. By using a robot to generate precise, ground-truth breathing patterns and an IR-LWS system with lock-in detection, the authors extract handcrafted features (peak-to-peak amplitude, breathing rate, effective spectral amplitude, and SNR) for multiclass and binary classification. Random Forest emerged as the top performer, achieving up to 96.75% accuracy at 0.5 m, and ensemble methods consistently outperformed a single decision tree across distances. The study highlights practical viability for noninvasive respiratory monitoring and fault data rejection, while noting limitations (synthetic data, lack of clinical validation) and proposing transfer learning and clinical trials for real-world adoption.

Abstract

Human respiratory rate and its pattern convey essential information about the physical and psychological states of the subject. Abnormal breathing can indicate fatal health issues leading to further diagnosis and treatment. Wireless light-wave sensing (LWS) using incoherent infrared light shows promise in safe, discreet, efficient, and non-invasive human breathing monitoring without raising privacy concerns. The respiration monitoring system needs to be trained on different types of breathing patterns to identify breathing anomalies.The system must also validate the collected data as a breathing waveform, discarding any faulty data caused by external interruption, user movement, or system malfunction. To address these needs, this study simulated normal and different types of abnormal respiration using a robot that mimics human breathing patterns. Then, time-series respiration data were collected using infrared light-wave sensing technology. Three machine learning algorithms, decision tree, random forest and XGBoost, were applied to detect breathing anomalies and faulty data. Model performances were evaluated through cross-validation, assessing classification accuracy, precision and recall scores. The random forest model achieved the highest classification accuracy of 96.75% with data collected at a 0.5m distance. In general, ensemble models like random forest and XGBoost performed better than a single model in classifying the data collected at multiple distances from the light-wave sensing setup.

Noncontact Respiratory Anomaly Detection Using Infrared Light-Wave Sensing

TL;DR

This work demonstrates noncontact respiration monitoring with infrared light-wave sensing (LWS) and machine-learning-based anomaly detection. By using a robot to generate precise, ground-truth breathing patterns and an IR-LWS system with lock-in detection, the authors extract handcrafted features (peak-to-peak amplitude, breathing rate, effective spectral amplitude, and SNR) for multiclass and binary classification. Random Forest emerged as the top performer, achieving up to 96.75% accuracy at 0.5 m, and ensemble methods consistently outperformed a single decision tree across distances. The study highlights practical viability for noninvasive respiratory monitoring and fault data rejection, while noting limitations (synthetic data, lack of clinical validation) and proposing transfer learning and clinical trials for real-world adoption.

Abstract

Human respiratory rate and its pattern convey essential information about the physical and psychological states of the subject. Abnormal breathing can indicate fatal health issues leading to further diagnosis and treatment. Wireless light-wave sensing (LWS) using incoherent infrared light shows promise in safe, discreet, efficient, and non-invasive human breathing monitoring without raising privacy concerns. The respiration monitoring system needs to be trained on different types of breathing patterns to identify breathing anomalies.The system must also validate the collected data as a breathing waveform, discarding any faulty data caused by external interruption, user movement, or system malfunction. To address these needs, this study simulated normal and different types of abnormal respiration using a robot that mimics human breathing patterns. Then, time-series respiration data were collected using infrared light-wave sensing technology. Three machine learning algorithms, decision tree, random forest and XGBoost, were applied to detect breathing anomalies and faulty data. Model performances were evaluated through cross-validation, assessing classification accuracy, precision and recall scores. The random forest model achieved the highest classification accuracy of 96.75% with data collected at a 0.5m distance. In general, ensemble models like random forest and XGBoost performed better than a single model in classifying the data collected at multiple distances from the light-wave sensing setup.
Paper Structure (24 sections, 8 equations, 15 figures, 2 tables)

This paper contains 24 sections, 8 equations, 15 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Human Breathing Patterns
  4. Respiration Sensing
  5. Theory of Operation
  6. Lock-in Detection
  7. Data Collection and Processing
  8. Robot-simulated Respiration
  9. LWS Hardware Setup
  10. Data Collection
  11. Data Processing
  12. Feature extraction
  13. Peak-to-peak Amplitude
  14. Breathing Rate
  15. Effective Spectral Amplitude
  16. ...and 9 more sections

Figures (15)

  • Figure 1: Proposed system model for respiratory anomaly detection using infrared (IR) light-wave sensing
  • Figure 2: The robot's chest-rib and motor assembly, operated by a Raspberry Pi.
  • Figure 3: The experimental setup diagram of the LWS system used for data collection.
  • Figure 4: The experimental setup showing the key components of the LWS system used for data collection.
  • Figure 5: Flow diagram of the approach towards the breathing anomaly detection problem.
  • ...and 10 more figures