Hand Flexibility Assessment System Study Based on Pneumatic Actuation and Sensing

Yuanqing Xia, Ting Gao, W.h. Wang, Yalan Wang, Siyu Chen, Yue Zhang, Xinyuan ChenRehabilitation 2025 · DOI

Abstract

Objective: To develop a hand flexibility assessment system based on pneumatic actuation and sensing, and validate its accuracy and reliability. Methods: A pneumatic-driven system consisting of actuation modules, circuit boards, data feedback, and human-computer interaction modules was designed. Air pressure serves as the quantitative indicator. Four impedance evaluation modes (mild to very severe) were established. Real-time air pressure data from soft actuators during finger flexion/extension were recorded for eight patients and compared against clinical scales (Brunnstrom and TAM) and physician assessments. Results: Healthy subjects showed consistent average pressure changes (~10 kPa flexion, ~6 kPa extension) with low standard deviations (0.13–0.40 kPa). Patient evaluations closely matched physician ratings, with the system detecting subtle 1 kPa differences. Conclusion: The system demonstrates high accuracy and reliability, offering a potential quantitative alternative to manual clinical assessments.

Introduction

Traditional hand flexibility assessment in rehabilitation often suffers from inter-rater variability due to reliance on subjective manual palpation or visual scales like the Brunnstrom stage. While technologies like computer vision or wearable IMU sensors exist, they are frequently limited by lighting conditions, occlusion, or signal drift. This study proposes a soft robotic approach using pneumatic actuation and sensing to provide a robust, objective, and repeatable measurement tool for tracking rehabilitation progress.

Methods

The developed system integrates four primary components: a pneumatic actuation module, a control circuit board, a data feedback module, and a human-computer interaction interface.

System architecture and data flow diagram showing the interaction between the pneumatic modules, sensors, and user interface.

The pneumatic module utilizes pumps and solenoid valves to drive soft actuators that assist or resist finger movement. High-precision pressure sensors monitor internal air pressure across four intensity levels (mild, moderate, severe, and very severe). Signal processing involves a moving average smoothing algorithm and second-order polynomial fitting to establish baseline trends. Hand flexibility is quantified by the variance between real-time pressure data and these fitted trend lines, correlating with established clinical metrics like the Brunnstrom stage and Total Active Motion (TAM).

Results and Discussion

Validation involved establishing a "healthy template" baseline and testing a clinical cohort of eight patients with various hand dysfunctions, including stroke and spinal cord injuries.

Real-time air pressure data graph illustrating step-wise pressure increases and impedance flexion test results.

The system demonstrated exceptional repeatability in healthy subjects, with standard deviations between 0.13 and 0.40 kPa across 10 trials. In clinical testing, the system’s automated grading showed a near-perfect match with assessments from three senior attending physicians. Notably, the system could distinguish between patients in similar recovery stages by detecting pressure differences as small as 1 kPa—a level of granularity difficult to achieve through manual observation.

Conclusion

The pneumatic-driven assessment system provides a scientific, quantitative basis for evaluating hand flexibility. While future work is needed to increase sample sizes and address environmental sensitivities such as ambient temperature, the current results indicate that this technology can reliably support clinicians in developing personalized rehabilitation plans.

Summary updated

Abstract

<bold>Objective</bold> To develop a hand flexibility assessment system based on pneumatic actuation and sensing, and validate its accuracy and reliability. The system aims to provide a scientific basis for solving the problem of hand flexibility assessment and developing hand rehabilitation training plans. <bold>Methods</bold> A pneumatic-driven assessment system consisting of pneumatic actuation modules, circuit boards, data feedback modules, and human-computer interaction modules is designed. The system uses air pressure as a quantitative indicator to assess hand flexibility. Based on the data collection needs of patients with varying degrees of hand dysfunction, four impedance evaluation modes with different intensities (mild, moderate, severe, and very severe) are designed. The soft actuator produces unilateral bending or relaxation through air inflation and deflation. During multiple finger flexions by the user, an opposing effect is generated between the user and the system. The air pressure inside the soft actuator is recorded in real time, providing data for hand flexibility assessment. Hand flexibility is quantified by combining the Brunnstrom hand function stage scale and the Total Active Motion (TAM) assessment method. Healthy subjects' test data are recorded and used as a template. The air pressure data of eight patients with hand dysfunction are collected for hand flexibility evaluation and compared with the results of three attending physicians from a rehabilitation center to verify the system’s accuracy in assessing hand flexibility. <bold>Results</bold> The average air pressure change was maintained around 10 kPa for healthy subjects performing group impedance flexion tests. In 10 repeated experiments, the standard deviations of test results for the four intensity levels (mild, moderate, severe, very severe) were 0.37535, 0.27961, 0.39354, and 0.32010, respectively. During group impedance extension tests, the average air pressure change was maintained around 6 kPa, with standard deviations for the four intensity levels of 0.40442, 0.26102, 0.22012, and 0.13050, respectively, in 10 repeated experiments. For patients 1 and 2, based on the Brunnstrom evaluation scale, their finger recovery stages were determined to be Stage I and Stage II, respectively. For patients 3 and 4, the air pressure change was around 1 kPa. Patient 3 demonstrated better finger flexibility than patient 4. Patients 5 to 8 were evaluated using the TAM method, and patient 8 was classified with good hand flexibility, while the others were rated as poor. <bold>Conclusion</bold> The system shows high accuracy and reliability in assessing hand flexibility and has the potential to replace rehabilitation physicians in quantitatively evaluating hand flexibility.
BibTeX
@inproceedings{xia2025hand,
  author = {Yuanqing Xia and Ting Gao and W.H. Wang and Yalan WANG and Siyu Chen and Yue Zhang and Xinyuan CHEN},
  title = {{Hand Flexibility Assessment System Study Based on Pneumatic Actuation and Sensing}},
  booktitle = {Rehabilitation},
  year = {2025},
  doi = {10.3724/sp.j.1329.2025.02012},
}
Figure