Development and Validation of Ambulosono: A Wearable Sensor for Bio-Feedback Rehabilitation Training

Taylor Chomiak, Abhijot S. Sidhu, Alexander Watts, Luke Su, Brian Barkley Graham, Joshua H. Wu, Suzanne Classen, Brian Falter, Bin HuSensors 2019 · DOI

TL;DR

This paper introduces and validates "Ambulosono," a new, iPod-independent wearable sensor system for bio-feedback rehabilitation training. It demonstrates comparable, and in some aspects superior, accuracy and test-retest reliability to the established iPod-based system, offering a more economical and accessible solution for clinical gait analysis.

What Problem Does This Paper Address?

Wearable sensor systems are increasingly recognized for their potential in managing chronic diseases and aiding rehabilitation. However, existing systems, like the one based on the iPod touch, can be costly and less accessible for widespread clinical use. The challenge is to develop a reliable, accurate, and affordable wearable sensor system that can effectively measure key functional parameters, such as gait, for therapeutic purposes. This paper aims to address this by developing and validating an iPod-independent version of their previously developed Ambulosono system.

Methods

The study developed a new Ambulosono sensor system that utilizes a Seeed Tiny BLE development board with an MPU-6050 motion processor. This system captures raw accelerometer data for motion detection and calibrated gyroscope quaternions for Euler angle conversion. The system was designed to be Bluetooth Low Energy (BLE) compatible, allowing it to connect to Android or iOS mobile devices.

For validation, the new Ambulosono sensor was compared against the functionality of an iPod touch (4th generation) sensor system. Both systems were attached to participants' thighs just above the knee using a high-performance thigh band. Participants performed a gait test in a 30.48m hallway. True gait parameters (step length, distance, velocity, cadence) were established using video analysis and floor markings. The sensor data was then processed using the "GaitReminder™ App," which employs algorithms to calculate gait parameters, accounting for limb length, angular excursion, and signal filtering. The accuracy was assessed by calculating the mean percentage error (MPE) between sensor-derived values and true values. Test-retest reliability was evaluated using the intra-class correlation coefficient (ICC).

Datasets

The study did not introduce new datasets for training. Instead, it used data collected from fifteen healthy adult male participants during a controlled gait test. The "true" gait parameters were established through video analysis and physical measurements of the testing environment. This served as the benchmark for evaluating the performance of the wearable sensors.

Key Insights and Ideas

  • iPod Independence is Key: The primary innovation is the development of an Ambulosono system that doesn't rely on an iPod touch, making it more adaptable to various mobile devices and potentially more cost-effective for clinical implementation.
  • Comparable, Yet Superior: While the new system performs comparably to the iPod-based version, it shows a tendency towards lower mean percentage errors in step length, distance, and velocity, suggesting potential for improved accuracy in specific metrics.
  • Reliability is High: Both the iPod-based and the new Ambulosono systems demonstrate strong test-retest reliability, indicating consistent measurements across repeated trials, which is crucial for clinical applications.
  • Economic Alternative: The findings suggest that Ambulosono can serve as a viable and more economical alternative to existing iPod-based systems, broadening accessibility for rehabilitation programs.

Key Results

  • Mean Percentage Error (MPE): For all measured gait parameters (step length, distance, velocity, cadence), the MPE was less than 10%.
  • Distance MPE: The Ambulosono sensor showed a statistically significant lower MPE for distance compared to the iPod sensor (t(14) = -2.33; p = 0.035).
  • Test-Retest Reliability (ICC):
    • iPod ICC ranged from 0.78 to 0.92.
  • Ambulosono ICC ranged from 0.82 to 0.92.
  • Both showed strong agreement (p < 0.001 for all parameters). The Ambulosono system performed slightly better in some reliability measures.

Limitations and Future Work

  • Small Sample Size: The study acknowledges that the relatively small sample size (n=15) limited their ability to statistically discern differences in MPE for step length and velocity between the sensor systems, despite observed average value differences.
  • Healthy Controls: The study focused on healthy participants. Future work needs to establish test-retest reliability for specific patient populations, as their gait parameters can exhibit greater variability.
  • Further Patient Studies: Ongoing research with Ambulosono across multiple sites indicates its potential for scaling up to high-volume clinical services and developing new paradigms for non-pharmacological care for various patient groups.

Summary updated

Abstract

Wearable technology-based measurement systems hold potential for the therapeutic and rehabilitation management of patients with various chronic diseases. The purpose of this study was to assess the accuracy and test⁻retest reliability of a new-generation wearable sensor-based system, dubbed Ambulosono, for bio-feedback training. The Ambulosono sensor system was cross-validated by comparing its functionality with the iPod touch (4th generation) sensor system. Fifteen participants underwent a gait test to measure various gait parameters while wearing both the iPod-based and Ambulosono sensors simultaneously. The physically measured values (i.e., the true values) of step length, distance traveled, velocity, and cadence were then compared to those obtained via the two-sensor systems using the same calculation algorithms. While the mean percentage error was <10% for all measured parameters, and the intra-class correlation coefficient revealed a high level of agreement between trials for both sensor systems, it was found that the Ambulosono sensor system outperformed the iPod-based system in some respects. The Ambulosono sensor system possessed both reliability and accuracy in obtaining gait parameter measurements, which suggests it can serve as an economical alternative to the iPod-based system that is currently used in various clinical rehabilitation programs.
BibTeX
@inproceedings{chomiak2019development,
  author = {Taylor Chomiak and Abhijot S. Sidhu and Alexander Watts and Luke Su and Brian Barkley Graham and Joshua H. Wu and Suzanne Classen and Brian Falter and Bin Hu},
  title = {{Development and Validation of Ambulosono: A Wearable Sensor for Bio-Feedback Rehabilitation Training}},
  booktitle = {Sensors},
  year = {2019},
  doi = {10.3390/s19030686},
}
Figure