Protocol for Psychometric Analysis and Retrofit of Device for Kinematic
Analysis of Gait After Stroke
Stroke; Gait Analysis; Assistive Technology; Wearable Devices.
Introduction: Gait deficits affect between 50–80% of all stroke
survivors. An alternative for analyzing gait kinematic parameters are
systems with inertial measurement units (IMUs), which can be wearable
and attached to different body segments, allowing analysis in the
individual's everyday environments. Objective: To develop a new version
(retrofit) of a portable device for the kinematic assessment of gait in post-
stroke individuals (DAMA), as well as a protocol for the psychometric
evaluation of this device. Methods: This is a study of the development
and measurement properties of DAMA, a device with two inertial sensors
to measure the angular displacement of the ankle during gait. A new
version of DAMA was developed based on previous tests conducted with
healthy individuals and post-stroke individuals, incorporating updates to
the structural components and the smartphone connection interface,
which now operates via Wi-Fi. The device now uses the ESP32 Firebeetle
V3 microcontroller, equipped with a 32-bit dual-core processor, and has
adopted the Inter-Integrated Circuit (I2C) protocol for communication
with the two inertial sensors: GY-BNO055 (internal sensor) and CJMCU-
055 (external sensor). The protocol for psychometric analysis was
developed to assess the criterion validity, test-retest reliability, and
usability (involving operational performance and user satisfaction) of
DAMA. Preliminary testing of the protocol and the new version of the
device was performed on an individual in the chronic post-stroke phase,
using the Qualisys Motion Capture System® as a reference measure. Data
collection with DAMA and the Qualisys® system was performed
simultaneously to analyze criterion validity and usability. To assess
reliability, two measurements were conducted: a test (first assessment)
and a retest after a seven-day interval. Regarding usability outcomes, an
author-developed form was used to assess the device's operational
performance. User satisfaction was measured using the Brazilian adapted
version of the Quebec User Evaluation of Satisfaction with Assistive
Technology (B-QUEST 2.0). As these are preliminary data, they were
presented in a descriptive manner. Results: A 66-year-old male patient
who had suffered a stroke 10 years ago completed the preliminary testing.
The preliminary performance of the device revealed differences in ankle
angle detection patterns, particularly in initial contact data compared to
Qualisys. In terms of operational performance, the device proved
satisfactory, especially in terms of battery life and real-time data
transmission. In terms of user satisfaction, the average total score for the
adapted B-QUEST 2.0 at both intervention times also indicated a
satisfactory level of usability. Conclusion: The developed device
demonstrated preliminary feasibility in terms of performance and
usability, showing potential for application in gait analysis. However,
adjustments are needed to automate the detection of gait events in order to
minimize possible biases resulting from manual interventions. In
addition, the hardware structure with external sensor connection should
be redesigned to improve adaptation to the volunteers' ankles and reduce
the risk of sensor calibration during use. Studies with larger and more
heterogeneous samples are needed to assess the reliability and validity of
the device in comparison to the gold standard, considering different gait patterns presented by post-stroke individuals.