Commercially available objective gait evaluation systems show good to excellent reliability in horses across hard-surface types and multiple operators.
Abstract: To evaluate the intrasystem reliability of an AI-based smartphone application Sleip (AI-SPA), the inertial measurement unit (IMU) system Equinosis Q Lameness Locator (LL), and the IMU system Equisym (ES) across 2 different hard surfaces and to assess the interoperator reliability for the 2 IMU systems. Unassigned: 8 mixed-breed research horses, aged 4 to 12 years, with varying degrees of gait asymmetries were utilized. To assess intrasystem reliability, horses were in-hand, straight-line trotted on either epoxy-coated concrete or rubber-coated concrete. Three trials were performed for LL and ES, while 6 concurrent trials were performed for AI-SPA. Interoperator reliability was assessed with 4 operators on the rubber-coated concrete only. Intraclass correlation coefficients and SE of measurements were calculated for 4 outcome parameters common to all systems as well as a variety of system-specific outcome parameters for LL and ES. Unassigned: Nearly all intraclass correlation coefficients indicated moderate to excellent intrasystem reliability (> 0.50) across the 3 systems (AI-SPA, 0.40 to 0.85; LL, 0.73 to 0.94; ES, 0.44 to 0.99) and were comparable between surfaces. Interoperator reliability was moderate to excellent for both LL (0.72 to 0.92) and ES (0.73 to 0.96). Standard error of measurement values remained relatively stable across study conditions. Unassigned: Despite inherent horse variability and minor differences in testing conditions, objective gait evaluation systems consistently produced measurements indicating moderate to excellent reliability and low measurement error. Unassigned: These findings indicate that observed changes in system output are unlikely to be impacted by operator or hard-surface effects, supporting clinical reliability of these systems in routine practice when utilizing manufacturers' instructions.
The Equine Research Bank provides access to a large database of publicly available scientific literature. Inclusion in the Research Bank does not imply endorsement of study methods or findings by Mad Barn.
This research summary has been generated with artificial intelligence and may contain errors and omissions. Refer to the original study to confirm details provided. Submit correction.
Overview
This study evaluated the reliability of three different horse gait analysis systems, including an AI-based smartphone app and two inertial measurement unit (IMU) systems, across different hard surfaces and multiple operators.
The goal was to determine how consistent these systems are when measuring horse gait and whether factors like surface type or who operates the device affect results.
Background and Purpose
Objective gait evaluation systems are tools designed to measure a horse’s movement and detect gait abnormalities, which are important for diagnosing lameness or other issues.
Reliability is critical for clinical use; the measurements should be consistent regardless of the testing environment or operator.
This study focused on three commercially available systems:
The AI-based smartphone application called Sleip (AI-SPA)
The Equinosis Q Lameness Locator (LL), an IMU-based system
The Equisym (ES), another IMU-based system
The researchers wanted to compare:
Intrasystem reliability — consistency of measurements within the same system across trials
Interoperator reliability — consistency between different users (operators) of the two IMU systems
Reliability across two different hard surfaces: epoxy-coated concrete and rubber-coated concrete
Methods
Subjects: 8 mixed-breed research horses aged 4 to 12 years with varying degrees of gait asymmetry were used.
Testing conditions:
Horses were trotted in-hand in a straight line on two types of hard surfaces.
For LL and ES systems, three trials per horse per surface were recorded.
For AI-SPA, six concurrent trials per horse per surface were recorded due to its smartphone-based nature.
Interoperator reliability was tested only on rubber-coated concrete with four different operators using the two IMU systems (LL and ES).
Data analysis:
Intraclass correlation coefficients (ICCs) measured reliability for each system’s parameters common across all devices.
Standard error of measurement (SEM) was calculated to quantify measurement error and variability.
Results
Intrasystem reliability:
All three systems showed moderate to excellent reliability (ICCs generally > 0.50), with:
AI-SPA ICCs between 0.40 and 0.85
LL ICCs between 0.73 and 0.94
ES ICCs between 0.44 and 0.99
Reliability was comparable across both hard surfaces.
Interoperator reliability:
For LL: Moderate to excellent reliability ranged from 0.72 to 0.92 across four operators.
For ES: Reliability ranged from 0.73 to 0.96 across operators.
Standard error of measurement:
SEM values remained relatively stable regardless of testing conditions, indicating low measurement variability.
Conclusions and Implications
Despite natural variations in horse gait and slight differences in testing conditions (e.g., surfaces and operators), the systems provided consistent and reliable gait measurements.
The moderate to excellent reliability and low measurement errors suggest that changes detected by these systems in clinical or research settings likely reflect true changes in gait rather than measurement noise.
The similar results across different hard surfaces and multiple operators indicate that:
Surface type and operator differences have minimal effect on measurement reliability when following manufacturers’ instructions.
These systems can be trusted for regular clinical use in assessing equine gait objectively.
The AI-SPA, being a smartphone app, shows promise as a convenient and reliable alternative to traditional IMU-based systems.
Cite This Article
APA
McPeek JL, Caudill A, Adam E, Carattini S, Jacob O, Page AE.
(2026).
Commercially available objective gait evaluation systems show good to excellent reliability in horses across hard-surface types and multiple operators.
J Am Vet Med Assoc, 1-11.
https://doi.org/10.2460/javma.26.06.0495