A consumer-grade IMU-based gait analysis workflow showed strongest inter-system agreement for sagittal-plane kinematics in transtibial amputees, while frontal- and transverse-plane measures and distal-joint variables should be interpreted cautiously.
Key Findings
Results
Repeated-trial consistency was high for hip kinematics across all three anatomical planes and for sagittal-plane knee and ankle kinematics in transtibial amputees.
Average-measure ICCs ranged from 0.945 to 0.996 for these measures.
Thirty-two individuals with unilateral transtibial amputation completed five overground level-walking trials.
Both the consumer-grade IMU (cIMU) and research-grade IMU (rIMU) systems were worn simultaneously with sensors mounted in standardized adjacent positions.
Repeated-trial consistency was assessed using intraclass correlation coefficients (ICCs) and intra-subject variability measures.
Results
Inter-workflow agreement between the consumer-grade and research-grade IMU systems was strongest for sagittal-plane measures.
Bland-Altman mean differences were within 2.5° for sagittal-plane knee and ankle measures.
Wider limits of agreement were observed for several frontal- and transverse-plane measures and distal-joint variables.
Statistical Parametric Mapping (SPM) identified significant waveform differences across multiple gait-cycle phases, particularly for frontal- and transverse-plane measures.
Paired comparisons showed that differences in discrete measures were concentrated in non-sagittal and distal-joint variables.
Methods
The consumer-grade IMU workflow used custom Python scripts to derive bilateral hip, knee, and ankle kinematics from segment-orientation outputs.
The workflow was described as 'transparent' and 'independently implemented,' distinguishing it from proprietary processing pipelines.
The research-grade IMU (rIMU) kinematics were generated using the system's proprietary processing software for comparison.
Kinematics were assessed across three anatomical planes: sagittal, frontal, and transverse.
Assessment methodology included SPM, Bland-Altman analysis, ICCs, intra-subject variability, and paired comparisons.
Results
Frontal- and transverse-plane kinematics and distal-joint measures showed lower inter-workflow agreement and should be interpreted cautiously when using the consumer-grade IMU workflow.
SPM identified significant waveform differences particularly for frontal- and transverse-plane measures across multiple gait-cycle phases.
Wider limits of agreement on Bland-Altman analysis were observed for frontal- and transverse-plane measures and distal-joint variables.
Paired comparisons confirmed that differences in discrete measures were concentrated in non-sagittal and distal-joint variables.
Assessment in transtibial prosthesis users is complicated by altered segment coordination, prosthetic component function, and limited anatomical landmarks.
Conclusions
The consumer-grade IMU workflow was identified as a potentially practical option for clinical settings where research-grade gait analysis systems are unavailable.
Clinical gait assessment in transtibial amputees is currently limited by the cost, technical demands, and space requirements of laboratory-based motion analysis.
Research-grade IMU systems provide a portable alternative but their hardware and software costs may restrict routine clinical use.
The cIMU workflow was described as most useful for 'repeated-trial assessment of broad sagittal-plane gait patterns.'
The workflow's transparency and use of openly implementable Python scripts were highlighted as features supporting accessibility.
What This Means
This research suggests that a low-cost, consumer-grade motion sensor (IMU) system combined with custom analysis scripts can provide clinically useful information about how people with below-knee amputations walk, particularly when measuring forward-and-backward (sagittal plane) movements of the hip, knee, and ankle. The study had 32 people with one-sided below-knee amputations walk on flat ground while wearing both an affordable consumer sensor system and a more expensive research-grade sensor system at the same time, then compared the results from both systems. The affordable system showed good agreement with the expensive system for measuring hip movements in all directions, and for knee and ankle movements in the sagittal plane, with measurement differences generally staying within 2.5 degrees.
However, the study also found meaningful limitations. Measurements of side-to-side (frontal plane) and rotational (transverse plane) movements, as well as measurements at the ankle joint, showed larger disagreements between the two systems and should be used with caution. This is particularly relevant for amputee gait assessment because prosthetic limbs move differently from natural limbs and have fewer reliable body landmarks for sensor placement, making accurate measurement more challenging.
This research matters because access to expensive motion analysis laboratories is limited in many clinical settings around the world, leaving clinicians without objective tools to track how well prosthetic fitting and rehabilitation are working. This study suggests that a transparent, openly reproducible workflow using affordable sensors could fill that gap for basic sagittal-plane gait monitoring, though clinicians should be aware that side-to-side and rotational movement data from such systems may not be reliable enough for detailed clinical decision-making.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Rattanakoch J, Limroongreungrat W, Guerra G, Manupibul U, Apidech H, Niamsang W, et al.. (2026). Assessment of IMU-derived lower-limb kinematics in transtibial amputees: A practical and accessible clinical workflow.. PloS one. https://doi.org/10.1371/journal.pone.0356786