Validation of Smart Insoles’ Fatigue Score, a Case Report




Journal Title

Journal ISSN

Volume Title



Background: Smart insoles give users valuable insights into their gait patterns, aiming to offer a cost-effective, reliable, accurate, and portable alternative to lab-based gait measurement using motion capture and force plates. Recent advancements have allowed IMU and pressure sensors (FSR) to be integrated directly into insoles, facilitating gait measurements outside of a lab environment. One promising application of smart insoles is to detect biomechanical changes in gait patterns and predict lower-extremity (LE) muscular fatigue. Reliable and accurate measurements of fatigue and recovery could play an essential role in injury prevention strategies and optimizing training intensity in various sports. Prior studies have used validated questionnaires such as the Rating of Fatigue (RoF) and Borg’s Rating of Perceived Exertion (RPE), which showed strong correlations with perceived fatigue and exertion ratings. Heart rate (HR) has also been frequently utilized among the objective measurements of fatigue. This case report aims to validate the fatigue score predicted by the smart insoles (LFS) against subjective and objective measures of fatigue such as RoF, RPE, and HR.

Case Information: A 24-year-old male without existing gait abnormalities participated in a pre-test/post-test design with a progression-to-fatigue protocol consisting of multiple sets of repeated exercises until fatigue. The baseline (pre-evaluation) assessment included a 20m fast-paced walk and 2 static movement jumps (SMJ), followed by 2 countermovement jumps (CMJ). RoF, RPE, and HR measurements were collected at baseline and after each fatiguing set. Each fatigue set consisted of 6x 10m shuttle sprints followed by 5 vertical jumps consecutively without rest. Following each fatigue set, participants underwent post-evaluations that included the same exercises as the pre-evaluation. This approach enabled the insoles to analyze changes in gait from baseline due to the accumulation of fatigue. A 30-minute recovery session was completed after the participant reached a RoF = 10. The recovery session consisted of the RoF, RPE, HR, and post-evaluation measurements every 2 minutes for the first 10 minutes, then every 5 minutes for the remaining 20 minutes.

Conclusion: The participant reached RoF = 10 after 6 repeated fatigue sets. Ground contact time (GCT), measured by the insoles, is one parameter that has been identified to inform the level of fatigue. In the fatiguing session, GCT was very strongly correlated with RoF and RPE (r = 0.971, p=.001); (r=0.986, p<0.001), respectively, but was not strongly correlated with HR (r= 0.543, p = 0.27). In the recovery session, GCT was strongly correlated with both RoF and HR (r =0.689, p=0.04); (r=0.723, p =0.028), respectively. RPE was a poor indicator of fatigue in the recovery session compared to RoF, GCT, and HR.GCT, along with other identified parameters, has the potential to be utilized in the development of the LFS. The LFS score analyzed changes in gait following fatigue and subsequently output a fatigue score for each set. Spearman’s rho and ICC were used to assess correlations and agreement between LFS, RoF, RPE, and HR. Data analysis was done using SPSS v29.0.