Abstract

Gait is impaired in musculoskeletal conditions, such as knee arthropathy. Gait analysis is used in clinical practice to inform diagnosis and to monitor disease progression or intervention response. However, clinical gait analysis relies on subjective visual observation of walking, as objective gait analysis has not been possible within clinical settings due to the expensive equipment, large-scale facilities, and highly trained staff required. Relatively low-cost wearable digital insoles may offer a solution to these challenges. In this work, we demonstrate how a digital insole measuring osteoarthritis-specific gait signatures yields similar results to the clinical gait-lab standard. To achieve this, we constructed a machine learning model, trained on force plate data collected in participants with knee arthropathy and controls. This model was highly predictive of force plate data from a validation set (area under the receiver operating characteristics curve [auROC] = 0.86; area under the precision-recall curve [auPR] = 0.90) and of a separate, independent digital insole dataset containing control and knee osteoarthritis subjects (auROC = 0.83; auPR = 0.86). After showing that digital insole derived gait characteristics are comparable to traditional gait measurements, we next showed that a single stride of raw sensor time series data could be accurately assigned to each subject, highlighting that individuals using digital insoles can be identified by their gait characteristics. This work provides a framework for a promising alternative to traditional clinical gait analysis methods, adds to the growing body of knowledge regarding wearable technology analytical pipelines, and supports clinical development of at-home gait assessments, with the potential to improve the ease, frequency, and depth of patient monitoring.

Data availability

Anonymized data and computer code to reproduce all figures will be made available as supplementary files to this manuscript. All relevant demographic and clinical information, all vGRF, derived gait characteristics, and raw sensor time series data, will be provided, in addition to R and Python scripts used to perform the analysis. Additionally, this information will be uploaded to a Regeneron GitHub account here: https://github.com/regeneron-mpds. This data will be made fully available prior to final publication of this manuscript. The GaitRec dataset is available online here: https://www.nature.com/articles/s41597-020-0481-z

Article and author information

Author details

  1. Matthew F Wipperman

    Precision Medicine, Regeneron, Tarrytown, United States
    For correspondence
    matthew.wipperman@regeneron.com
    Competing interests
    Matthew F Wipperman, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-1436-3366
  2. Allen Z Lin

    Molecular Profiling and Data Science, Regeneron, Tarrytown, United States
    Competing interests
    Allen Z Lin, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  3. Kaitlyn M Gayvert

    Molecular Profiling and Data Science, Regeneron, Tarrytown, United States
    Competing interests
    Kaitlyn M Gayvert, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  4. Benjamin Lahner

    Precision Medicine, Regeneron, Tarrytown, United States
    Competing interests
    Benjamin Lahner, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  5. Selin Somersan-Karakaya

    Early Clinical Development and Experimental Sciences, Regeneron, Tarrytown, United States
    Competing interests
    Selin Somersan-Karakaya, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  6. Xuefang Wu

    Clinical Outcomes Assessment and Patient Innovation, Regeneron, Tarrytown, United States
    Competing interests
    Xuefang Wu, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  7. Joseph Im

    Clinical Outcomes Assessment and Patient Innovation, Regeneron, Tarrytown, United States
    Competing interests
    Joseph Im, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  8. Minji Lee

    Molecular Profiling and Data Science, Regeneron, Tarrytown, United States
    Competing interests
    Minji Lee, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  9. Bharatkumar Koyani

    Clinical Outcomes Assessment and Patient Innovation, Regeneron, Tarrytown, United States
    Competing interests
    Bharatkumar Koyani, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  10. Ian Setliff

    Molecular Profiling and Data Science, Regeneron, Tarrytown, United States
    Competing interests
    Ian Setliff, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  11. Malika Thakur

    Clinical Outcomes Assessment and Patient Innovation, Regeneron, Tarrytown, United States
    Competing interests
    Malika Thakur, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  12. Daoyu Duan

    Precision Medicine, Regeneron, Tarrytown, United States
    Competing interests
    Daoyu Duan, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  13. Aurora Breazna

    Biostatistics and Data Management, Regeneron, Tarrytown, United States
    Competing interests
    Aurora Breazna, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  14. Fang Wang

    Precision Medicine, Regeneron, Tarrytown, United States
    Competing interests
    Fang Wang, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  15. Wei Keat Lim

    Molecular Profiling and Data Science, Regeneron, Tarrytown, United States
    Competing interests
    Wei Keat Lim, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-6226-2570
  16. Gabor Halasz

    Molecular Profiling and Data Science, Regeneron, Tarrytown, United States
    Competing interests
    Gabor Halasz, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  17. Jacek Urbanek

    Biostatistics and Data Management, Regeneron, Tarrytown, United States
    Competing interests
    Jacek Urbanek, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  18. Yamini Patel

    General Medicine, Regeneron, Tarrytown, United States
    Competing interests
    Yamini Patel, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  19. Gurinder S Atwal

    Molecular Profiling and Data Science, Regeneron, Tarrytown, United States
    Competing interests
    Gurinder S Atwal, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  20. Jennifer D Hamilton

    Precision Medicine, Regeneron, Tarrytown, United States
    Competing interests
    Jennifer D Hamilton, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  21. Samuel Stuart

    Precision Medicine, Regeneron, Tarrytown, United States
    Competing interests
    Samuel Stuart, Employee and stockholder of Regeneron Pharmaceuticals, Inc..
  22. Oren Levy

    Early Clinical Development and Experimental Sciences, Regeneron, Tarrytown, United States
    Competing interests
    Oren Levy, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  23. Andreja Avbersek

    Early Clinical Development and Experimental Sciences, Regeneron, Tarrytown, United States
    Competing interests
    Andreja Avbersek, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  24. Rinol Alaj

    Clinical Outcomes Assessment and Patient Innovation, Regeneron, Tarrytown, United States
    For correspondence
    rinol.alaj@regeneron.com
    Competing interests
    Rinol Alaj, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  25. Sara C Hamon

    Precision Medicine, Regeneron, Tarrytown, United States
    For correspondence
    sara.hamon@regeneron.com
    Competing interests
    Sara C Hamon, Employee and shareholder of Regeneron Pharmaceuticals, Inc..
  26. Olivier Harari

    Early Clinical Development and Experimental Sciences, Regeneron, Tarrytown, United States
    For correspondence
    olivier.harari@regeneron.com
    Competing interests
    Olivier Harari, Employee and shareholder of Regeneron Pharmaceuticals, Inc..

Funding

Regeneron Pharmaceuticals (N/A)

  • Matthew F Wipperman
  • Allen Z Lin
  • Kaitlyn M Gayvert
  • Benjamin Lahner
  • Selin Somersan-Karakaya
  • Xuefang Wu
  • Joseph Im
  • Minji Lee
  • Bharatkumar Koyani
  • Ian Setliff
  • Malika Thakur
  • Daoyu Duan
  • Aurora Breazna
  • Fang Wang
  • Wei Keat Lim
  • Gabor Halasz
  • Jacek Urbanek
  • Yamini Patel
  • Gurinder S Atwal
  • Jennifer D Hamilton
  • Oren Levy
  • Andreja Avbersek
  • Rinol Alaj
  • Sara C Hamon
  • Olivier Harari

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Ethics

Human subjects: For the pilot study, participants were recruited internally within the Regeneron facility located in Tarrytown, NY, USA, and were provided written informed consent prior to participation. The study was considered exempt research under the Common Rule (45 CFR Sec 46.104). The R5069-OA-1849 study protocol received Institutional Review Board and ethics committee approvals from Moldova Medicines and Medical Device Agency and National Ethics Committee for Moldova, and the Western Institutional Review Board.

Copyright

© 2024, Wipperman et al.

This article is distributed under the terms of the Creative Commons Attribution License permitting unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,053
    views
  • 171
    downloads
  • 1
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Matthew F Wipperman
  2. Allen Z Lin
  3. Kaitlyn M Gayvert
  4. Benjamin Lahner
  5. Selin Somersan-Karakaya
  6. Xuefang Wu
  7. Joseph Im
  8. Minji Lee
  9. Bharatkumar Koyani
  10. Ian Setliff
  11. Malika Thakur
  12. Daoyu Duan
  13. Aurora Breazna
  14. Fang Wang
  15. Wei Keat Lim
  16. Gabor Halasz
  17. Jacek Urbanek
  18. Yamini Patel
  19. Gurinder S Atwal
  20. Jennifer D Hamilton
  21. Samuel Stuart
  22. Oren Levy
  23. Andreja Avbersek
  24. Rinol Alaj
  25. Sara C Hamon
  26. Olivier Harari
(2024)
Digital wearable insole-based identification of knee arthropathies and gait signatures using machine learning
eLife 13:e86132.
https://doi.org/10.7554/eLife.86132

Share this article

https://doi.org/10.7554/eLife.86132

Further reading

    1. Medicine
    2. Neuroscience
    Joanna Kosinska, Julian C Assmann ... Markus Schwaninger
    Research Article

    Monomethyl fumarate (MMF) and its prodrug dimethyl fumarate (DMF) are currently the most widely used agents for the treatment of multiple sclerosis (MS). However, not all patients benefit from DMF. We hypothesized that the variable response of patients may be due to their diet. In support of this hypothesis, mice subjected to experimental autoimmune encephalomyelitis (EAE), a model of MS, did not benefit from DMF treatment when fed a lauric acid-rich (LA) diet. Mice on normal chow (NC) diet, in contrast, and even more so mice on high-fiber (HFb) diet showed the expected protective DMF effect. DMF lacked efficacy in the LA diet-fed group despite similar resorption and preserved effects on plasma lipids. When mice were fed the permissive HFb diet, the protective effect of DMF treatment depended on hydroxycarboxylic receptor 2 (HCAR2) which is highly expressed in neutrophil granulocytes. Indeed, deletion of Hcar2 in neutrophils abrogated DMF protective effects in EAE. Diet had a profound effect on the transcriptional profile of neutrophils and modulated their response to MMF. In summary, DMF required HCAR2 on neutrophils as well as permissive dietary effects for its therapeutic action. Translating the dietary intervention into the clinic may improve MS therapy.

    1. Medicine
    Hyun Beom Song, Laura Campello ... Anand Swaroop
    Research Advance

    Inherited retinal degenerations (IRDs) constitute a group of clinically and genetically diverse vision-impairing disorders. Retinitis pigmentosa (RP), the most common form of IRD, is characterized by gradual dysfunction and degeneration of rod photoreceptors, followed by the loss of cone photoreceptors. Recently, we identified reserpine as a lead molecule for maintaining rod survival in mouse and human retinal organoids as well as in the rd16 mouse, which phenocopy Leber congenital amaurosis caused by mutations in the cilia-centrosomal gene CEP290 (Chen et al., 2023). Here, we show the therapeutic potential of reserpine in a rhodopsin P23H rat model of autosomal dominant RP. At postnatal day (P) 68, when males and females are analyzed together, the reserpine-treated rats exhibit higher rod-derived scotopic b-wave amplitudes compared to the controls with little or no change in scotopic a-wave or cone-derived photopic b-wave. Interestingly, the reserpine-treated female rats display enhanced scotopic a- and b-waves and photopic b-wave responses at P68, along with a better contrast threshold and increased outer nuclear layer thickness. The female rats demonstrate better preservation of both rod and cone photoreceptors following reserpine treatment. Retinal transcriptome analysis reveals sex-specific responses to reserpine, with significant upregulation of phototransduction genes and proteostasis-related pathways, and notably, genes associated with stress response. This study builds upon our previously reported results reaffirming the potential of reserpine for gene-agnostic treatment of IRDs and emphasizes the importance of biological sex in retinal disease research and therapy development.