Towards the Monitoring of Functional Status in a Free-Living Environment for People with Hip or Knee Osteoarthritis: Design and Evaluation of the JOLO Blended Care App
<p>Screenshots of the patient interfaces (from <b>left</b> to <b>right</b>) activities that can be viewed in a day (=dag), week (=week) and month (=maand) overview. The app provides the total minutes active during that day (= minuten actief vandaag). The functional status (=Gewrichtsgebruik) gives a similar overview as the activities in a day, week and month. The goals (=doelen), in the third panel, show the goals that are set for that week. If a goal has been specified for an activity, the app shows how close the user is to reaching it. In this example, the goal is to walk (=wandelen) for 15 min, four times a week. These interfaces will be used on the smartphone app by the osteoarthritis (OA) patients.</p> "> Figure 2
<p>Screenshots of the (1) activities, (2) functional status, and (3) goal screen of the clinical dashboard for the health-care professionals. The activities are displayed in stacked bar graphs per day for walking (=wandelen), cycling (=fietsen), sit-to-stand (ZitNaarStand), stand-to-sit (=StaanNaarZit), and running (=joggen). The second graph shows the corresponding functional status for that time period. The third overview shows the goals (=Doelen) for each time period. The considered time periods are 6 weeks (=weken), 3 months (=maanden), 6 months (=maanden), and 1 year (=jaar).</p> "> Figure 3
<p>Schematic overview of the JOLO (Joint Load) architecture. The JOLO database (DB JOLO) can be considered as the central component of the architecture. Using API calls, it provides information to both the app and the clinical dashboard. The app regularly uploads data to Amazon S3 and the backend server. Uploaded data is asynchronously fetched, processed by the activity toolbox, postprocessed and stored in the JOLO DB. App and clinical dashboard authorization is managed by a separate database (DB Dharma).</p> "> Figure 4
<p>Accuracy and percentage of non-replaced data for different <math display="inline"><semantics> <mi>δ</mi> </semantics></math> thresholds of postprocessing step 1.</p> "> Figure 5
<p>Data collection and JOLO refinement flow divided into two phases, each with two iterations. From each usability test session, invaluable information was derived that was implemented in the design of the prototype that followed it.</p> "> Figure 6
<p>Raincloud plot [<a href="#B42-sensors-20-06967" class="html-bibr">42</a>,<a href="#B43-sensors-20-06967" class="html-bibr">43</a>] of the System Usability Scale (SUS) for the first and second iteration of the low-fidelity prototype for the smartphone app and the clinical dashboard. The maximum score is 100.</p> "> Figure 7
<p>Radar plot of the Unified Theory of Acceptance and Use of Technology (UTAUT) scores of the high-fidelity prototype of the smartphone app (plot on the <b>left</b>) and the clinical dashboard (plot on the <b>right</b>). A score of 3.5 is the neutral score; thus, everything above the 3.5 is a positive score (except anxiety, which should be below 3.5). The acronyms in the plots are as follows: PE = performance expectancy, AT = attitude towards using technology, ANX = anxiety, EE = effort expectancy, SI = social influence, FC = facilitating conditions, BI = behavioral intent and SE = self-efficacy.</p> "> Figure 8
<p>Radar plot of the Ease of Use Questionnaire (USE) scores of the high-fidelity prototype of the smartphone app (plot on the <b>left</b>) and the clinical dashboard (plot on the <b>right</b>). A score of 3.5 is a neutral score; thus, everything above the 3.5 is considered a positive score.</p> "> Figure 9
<p>Raincloud plot [<a href="#B42-sensors-20-06967" class="html-bibr">42</a>,<a href="#B43-sensors-20-06967" class="html-bibr">43</a>] of the Credibility/Expectancy Questionnaire (CEQ) scores of the high-fidelity prototype of the smartphone app. The maximum score of this questionnaire is 27, and everything above 13.5 is considered to be a positive score.</p> ">
Abstract
:1. Introduction
2. Materials and Methods
2.1. System Design and Developments
2.1.1. Module 1: JOLO Framework: Technologies and Architecture
- PostGreSQL: Processed patient activities are stored in the JOLO PostGreSQL database. Patient movement data are pseudonymized and stored in an encrypted way in the JOLO database compliant to GDPR rules and guidelines. A separate database contains patient personal information and is only accessible by the authorized researchers in a virtual private cloud.
- Laravel: The postprocessing backend and clinical dashboard are built using the Laravel PHP framework, which allows for a modular and flexible state-of-the-art approach to software architecture.
- Python: The toolbox that infers patient activities from sensor data is written as a Python script.
- Kotlin: The JOLO smartphone app for Android is written in Kotlin.
2.1.2. Module 2: Activity Recognition
- Preprocessing: Instead of using the raw tri-axial accelerometer data (, , ), we took the absolute value along each axis (, , ) in order to correct for different orientations of the mobile phone. Consequently, the user could position the phone in any orientation in the hip bag.
- Segmentation: In order to detect the activity performed at each second, we followed a sliding window approach with a step size of one second (50 samples) and a size of three seconds (150 samples). We labeled each window with the activity performed by the participant.
- Feature extraction: We transformed the preprocessed accelerometer measurements (, , ) to a feature representation to reduce the dimensionality of the data and reduce the risk of overfitting. Using the tsfresh Python package [28], we extracted the same set of 794 features from each acceleration signal. Each feature summarized the 150 samples of one signal as a single value. Examples of such features are the mean acceleration, the number of zero crossings, Fourier transform coefficients, etc. Using the same package, we selected a set of 98 relevant features from the features.
- Training: In order to correlate the features to the activity, we trained a gradient-boosted decision tree ensemble model that predicted the activity based on the features extracted for one window. We used the CatBoost Python package [29] to train the model.
- Instead of simply predicting the activity of each window, we also predicted the probability that a person was performing each of the seven activities. The highest probability (i.e., the probability corresponding to the predicted activity) provided an indication about how certain the model was. When this probability was lower than a threshold (), we replaced the predicted activity with an “unknown” activity label. For example, if the model predicted “descending stairs” for a certain time window but the probability associated with “descending stairs” was lower than , we replaced the predicted label “descending stairs” with the label “unknown”. The model may be uncertain in its predictions when the user performs other activities beyond the seven activities in our dataset, and when the user transitions from one activity to another. Replacing the prediction with “unknown” can reduce the number of incorrect predictions in these cases. To determine , we evaluated the accuracy as well as the percentage of data that was not replaced with an “unknown” label. Figure 4 shows the results for this experiment. Based on these results, we set to 0.5, where the accuracy improved slightly (from 72.82% to 74.43%) while 96% of the data were still not replaced with an “unknown” label. Taking a higher value for would improve the accuracy further, but this would also increase the number of predictions that are replaced by an “unknown” label. This is not desirable, as we may risk large chunks of activities being replaced by “unknown” even though they were detected correctly, albeit with a small probability.
- For sitting and standing, the model only detects sit–stand transitions. Thus, periods of sitting and standing are always misclassified as one of the seven activities. However, it is easy to detect these periods, as sitting and standing are the only activities in which the phone is not moving. We add a rule that classifies a window as “sitting or standing” when the variance of the acceleration is low (smaller than 0.1) and applies the learned model otherwise. Note that we cannot discriminate between sitting and standing, as the user is allowed to wear the phone in any orientation.
2.1.3. Module 3: Functional Status
2.1.4. Module 4: Personalized Goals
2.2. Study Design
2.3. Study Sample
2.4. Data Collection
2.4.1. Low-Fidelity Prototype
2.4.2. High-Fidelity Prototype
2.4.3. Questionnaire Scoring
3. Results
3.1. Phase I: Low-Fidelity Prototype
3.2. Phase II: High-Fidelity Prototype
4. Discussion
5. Conclusions
Author Contributions
Funding
Acknowledgments
Conflicts of Interest
Abbreviations
OA | Osteoarthritis |
SUS | System Usability Scale |
USE | Usability, Satisfaction and Ease of Use |
UTAUT | Unified Theory of Acceptance and Use of Technology |
CEQ | Credibility/Expectancy Questionnaire |
Appendix A
Appendix A.1
Appendix A.2
Appendix B
- You want to know if you have reached your goals on the 1st of June. Which information do you need, and where would you find that information?
- You want to know on which day you have been most active and which activities you have performed. Where can you find this information?
- You want to know on which day you have been closest to the joint loading goals. Which information do you need, and where would you find that information?
- Have you reached your goal joint loading in the month of June?
- You want to send a message to your therapist. Is this possible in this application?
- You want to know how your joint loading evolved over the month of June. Which steps do you need to follow to retrieve this information?
Appendix C
Smartphone App | Clinical Dashboard | ||
---|---|---|---|
Prototype I | Prototype II | Prototype I | |
Median [IQR] | Median [IQR] | Median [IQR] | |
Performance expectancy | 6.00 [4.75–6.25] | 2.75 [2.50–3.75] | 2.75 [2.75–4.25] |
Attitude towards using technology | 6.00 [5.40–6.20] | 5.00 [2.60–6.00] | 5.00 [3.20–5.20] |
Anxiety | 3.75 [3.25–4.50] | 1.75 [1.00–1.75] | 1.00 [1.00–1.25] |
Effort expectancy | 5.00 [4.75–5.75] | 4.50 [4.00–5.25] | 5.75 [5.50–6.25] |
Social influence | 5.75 [5.00–5.75] | 3.25 [3.00–3.25] | 2.00 [1.33–2.33] |
Facilitating conditions | 4.75 [4.75–5.75] | 5.25 [3.50–5.50] | 5.67 [5.33–6.33] |
Behavioral intent | 5.33 [5.00–6.00] | 2.67 [2.00–4.33] | 2.33 [2.33–3.33] |
Self-efficacy | 4.00 [3.75–4.50] | 4.25 [3.75–5.00] | 4.75 [4.00–5.25] |
Smartphone App | Clinical Dashboard | ||
---|---|---|---|
Prototype I | Prototype II | Prototype I | |
Median [IQR] | Median [IQR] | Median [IQR] | |
Usefulness | 5.25 [5.25–6.00] | 3.50 [2.25–4.37] | 3.00 [2.87–3.87] |
Ease of use | 5.18 [4.82–5.27] | 3.91 [3.00–5.27] | 5.09 [4.91–5.64] |
Ease of learning | 5.00 [4.50–5.50] | 4.00 [3.00–6.00] | 6.25 [6.00–6.50] |
Satisfaction | 5.14 [5.00–5.86] | 3.14 [2.00–5.00] | 3.29 [2.71–4.00] |
References
- Woolf, A.D.; Pfleger, B. Burden of major musculoskeletal conditions. Bull. World Health Organ. 2003, 81, 646–656. [Google Scholar] [PubMed]
- Felson, D.T. Osteoarthritis as a disease of mechanics. Osteoarthr. Cartil. 2013, 21, 10–15. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Maly, M.R. Abnormal and cumulative loading in knee osteoarthritis. Curr. Opin. Rheumatol. 2008, 20, 547–552. [Google Scholar] [CrossRef] [PubMed]
- Saxby, D.; Lloyd, D. Osteoarthritis year in review 2016: Mechanics. Osteoarthr. Cartil. 2017, 25, 190–198. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Miyazaki, T.; Wada, M.; Kawahara, H.; Sato, M.; Baba, H.; Shimada, S. Dynamic load at baseline can predict radiographic disease progression in medial compartment knee osteoarthritis. Ann. Rheum. Dis. 2002, 61, 617–622. [Google Scholar] [CrossRef] [PubMed]
- Tateuchi, H.; Koyama, Y.; Akiyama, H.; Goto, K.; So, K.; Kuroda, Y.; Ichihashi, N. Daily cumulative hip moment is associated with radiographic progression of secondary hip osteoarthritis. Osteoarthr. Cartil. 2017, 25, 1291–1298. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Musumeci, G. The effect of mechanical loading on articular cartilage. J. Funct. Morphol. Kinesiol. 2016, 1, 154–161. [Google Scholar] [CrossRef] [Green Version]
- McAlindon, T.E.; Bannuru, R.; Sullivan, M.; Arden, N.; Berenbaum, F.; Bierma-Zeinstra, S.; Hawker, G.; Henrotin, Y.; Hunter, D.; Kawaguchi, H.; et al. OARSI guidelines for the non-surgical management of knee osteoarthritis. Osteoarthr. Cartil. 2014, 22, 363–388. [Google Scholar] [CrossRef] [Green Version]
- Harris, W.H.; Sledge, C.B. Total hip and total knee replacement. N. Engl. J. Med. 1990, 323, 725–731. [Google Scholar] [CrossRef]
- Ernstgård, A.; PirouziFard, M.; Thorstensson, C.A. Health enhancing physical activity in patients with hip or knee osteoarthritis-an observational intervention study. BMC Musculoskelet. Disord. 2017, 18, 42. [Google Scholar] [CrossRef] [Green Version]
- Bennell, K.; Hinman, R.S.; Wrigley, T.V.; Creaby, M.W.; Hodges, P. Exercise and osteoarthritis: Cause and effects. Compr. Physiol. 2011, 1, 1943–2008. [Google Scholar] [PubMed]
- Zampogna, B.; Papalia, R.; Papalia, G.F.; Campi, S.; Vasta, S.; Vorini, F.; Fossati, C.; Torre, G.; Denaro, V. The Role of Physical Activity as Conservative Treatment for Hip and Knee Osteoarthritis in Older People: A Systematic Review and Meta-Analysis. J. Clin. Med. 2020, 9, 1167. [Google Scholar] [CrossRef] [PubMed]
- Van Rossom, S.; Smith, C.R.; Thelen, D.G.; Vanwanseele, B.; Van Assche, D.; Jonkers, I. Knee joint loading in healthy adults during functional exercises: Implications for rehabilitation guidelines. J. Orthop. Sports Phys. Ther. 2018, 48, 162–173. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Mündermann, A.; Dyrby, C.O.; D’Lima, D.D.; Colwell, C.W., Jr.; Andriacchi, T.P. In vivo knee loading characteristics during activities of daily living as measured by an instrumented total knee replacement. J. Orthop. Res. 2008, 26, 1167–1172. [Google Scholar] [CrossRef]
- Corlan, A.D. Medline Trend: Automated Yearly Statistics of PubMed Results for Any Query. Available online: http://dan.corlan.net/medline-trend.html (accessed on 22 June 2020).
- Deloitte’s 2019 Global Mobile Consumer Survey. Available online: https://www2.deloitte.com/us/en/insights/industry/telecommunications/global-mobile-consumer-survey.html (accessed on 8 October 2020).
- Palestra, G.; Rebiai, M.; Courtial, E.; Koutsouris, D. Evaluation of a Rehabilitation System for the Elderly in a Day Care Center. Information 2019, 10, 3. [Google Scholar] [CrossRef] [Green Version]
- Romeo, A.; Edney, S.; Plotnikoff, R.; Curtis, R.; Ryan, J.; Sanders, I.; Crozier, A.; Maher, C. Can smartphone apps increase physical activity? Systematic review and meta-analysis. J. Med. Internet Res. 2019, 21, e12053. [Google Scholar] [CrossRef]
- Middelweerd, A.; Mollee, J.S.; van der Wal, C.N.; Brug, J.; Te Velde, S.J. Apps to promote physical activity among adults: A review and content analysis. Int. J. Behav. Nutr. Phys. Act. 2014, 11, 97. [Google Scholar] [CrossRef] [Green Version]
- Danbjørg, D.B.; Villadsen, A.; Gill, E.; Rothmann, M.J.; Clemensen, J. Usage of an exercise app in the care for people with osteoarthritis: User-driven exploratory study. JMIR mHealth uHealth 2018, 6, e11. [Google Scholar] [CrossRef]
- Bandura, A. Health promotion by social cognitive means. Health Educ. Behav. 2004, 31, 143–164. [Google Scholar] [CrossRef]
- Bennell, K.L.; Dobson, F.; Hinman, R.S. Exercise in osteoarthritis: Moving from prescription to adherence. Best Pract. Res. Clin. Rheumatol. 2014, 28, 93–117. [Google Scholar] [CrossRef]
- Revenäs, Å.; Opava, C.H.; Martin, C.; Demmelmaier, I.; Keller, C.; Åsenlöf, P. Development of a web-based and mobile app to support physical activity in individuals with rheumatoid arthritis: Results from the second step of a co-design process. JMIR Res. Protoc. 2015, 4, e22. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Bossen, D.; Buskermolen, M.; Veenhof, C.; de Bakker, D.; Dekker, J. Adherence to a web-based physical activity intervention for patients with knee and/or hip osteoarthritis: A mixed method study. J. Med. Internet Res. 2013, 15, e223. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Kloek, C.J.; Bossen, D.; Veenhof, C.; van Dongen, J.M.; Dekker, J.; de Bakker, D.H. Effectiveness and cost-effectiveness of a blended exercise intervention for patients with hip and/or knee osteoarthritis: Study protocol of a randomized controlled trial. BMC Musculoskelet. Disord. 2014, 15, 269. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Anguita, D.; Ghio, A.; Oneto, L.; Parra, X.; Reyes-Ortiz, J.L. A public domain dataset for human activity recognition using smartphones. Esann 2013, 3, 3. [Google Scholar]
- Shoaib, M.; Bosch, S.; Incel, O.D.; Scholten, H.; Havinga, P.J. A survey of online activity recognition using mobile phones. Sensors 2015, 15, 2059–2085. [Google Scholar] [CrossRef]
- Christ, M.; Braun, N.; Neuffer, J.; Kempa-Liehr, A.W. Time series feature extraction on basis of scalable hypothesis tests (tsfresh–a python package). Neurocomputing 2018, 307, 72–77. [Google Scholar] [CrossRef]
- Prokhorenkova, L.; Gusev, G.; Vorobev, A.; Dorogush, A.V.; Gulin, A. CatBoost: Unbiased boosting with categorical features. In Proceedings of the Advances in neural information processing systems, Montréal, QC, Canada, 3–8 December 2018; pp. 6638–6648. [Google Scholar]
- Ichinoseki-Sekine, N.; Kuwae, Y.; Higashi, Y.; Fujimoto, T.; Sekine, M.; Tamura, T. Improving the accuracy of pedometer used by the elderly with the FFT algorithm. Med. Sci. Sports Exerc. 2006, 38, 1674–1681. [Google Scholar] [CrossRef]
- Emmerzaal, J.; van der Straaten, R.; Wesseling, M.; Bruijnes, A.; Corten, K.; De Baets, L.; Jonkers, I.; Timmermans, A.; Vanwanseele, B. Hip cf impulse during stair descent best discriminates between healthy controls and hip osteoarthritis patients. In Proceedings of the International Society of Biomechanics, Calgary, AB, Canada, 31 July–4 August 2019. [Google Scholar]
- Delp, S.L.; Anderson, F.C.; Arnold, A.S.; Loan, P.; Habib, A.; John, C.T.; Guendelman, E.; Thelen, D.G. OpenSim: Open-source software to create and analyze dynamic simulations of movement. IEEE Trans. Biomed. Eng. 2007, 54, 1940–1950. [Google Scholar] [CrossRef] [Green Version]
- Wesseling, M.; De Groote, F.; Meyer, C.; Corten, K.; Simon, J.P.; Desloovere, K.; Jonkers, I. Gait alterations to effectively reduce hip contact forces. J. Orthop. Res. 2015, 33, 1094–1102. [Google Scholar] [CrossRef]
- WHO. Global Recommendations on Physical Activity for Health. Available online: https://apps.who.int/iris/bitstream/handle/10665/44399/9789241599979_eng.pdf;jsessionid=D74E29638A8D93ACAAF90CA1C8D671F1?sequence=1 (accessed on 14 July 2020).
- Lewis, J.R. Usability: Lessons learned and yet to be learned. Int. J. Hum. Comput. Interact. 2014, 30, 663–684. [Google Scholar] [CrossRef]
- Brooke, J. SUS-A quick and dirty usability scale. Usability Eval. Ind. 1996, 189, 4–7. [Google Scholar]
- Lund, A.M. Measuring usability with the use questionnaire12. Usability Interface 2001, 8, 3–6. [Google Scholar]
- Venkatesh, V.; Morris, M.G.; Davis, G.B.; Davis, F.D. User acceptance of information technology: Toward a unified view. MIS Q. 2003, 27, 425–478. [Google Scholar] [CrossRef] [Green Version]
- Devilly, G.J.; Borkovec, T.D. Psychometric properties of the credibility/expectancy questionnaire. J. Behav. Ther. Exp. Psychiatry 2000, 31, 73–86. [Google Scholar] [CrossRef]
- Brooke, J. SUS: A retrospective. J. Usability Stud. 2013, 8, 29–40. [Google Scholar]
- Lewis, J.R. The system usability scale: Past, present, and future. Int. J. Hum. Comput. Interact. 2018, 34, 577–590. [Google Scholar] [CrossRef]
- Allen, M.; Poggiali, D.; Whitaker, K.; Marshall, T.R.; Kievit, R.A. Raincloud plots: A multi-platform tool for robust data visualization. [version 1; peer review: 2 approved]. Wellcome Open Res. 2019, 4. [Google Scholar] [CrossRef] [Green Version]
- Whitaker, K.; TomRhysMarshall; van Mourik, T.; Martinez, P.A.; Poggiali, D.; Ye, H.; MariusKlug. RainCloudPlots/RainCloudPlots: WellcomeOpenResearch. Wellcome Open Res. 2019. [Google Scholar] [CrossRef]
- Nielsen, J. Iterative user-interface design. Computer 1993, 26, 32–41. [Google Scholar] [CrossRef]
- Harrison, S.; Stadler, M.; Ismail, K.; Amiel, S.; Herrmann-Werner, A. Are Patients with Diabetes Mellitus Satisfied with Technologies Used to Assist with Diabetes Management and Coping? A Structured Review. Diabetes Technol. Ther. 2014, 16, 771–783. [Google Scholar] [CrossRef]
- Mills, K.; Hunt, M.A.; Ferber, R. Biomechanical deviations during level walking associated with knee osteoarthritis: A systematic review and meta-analysis. Arthritis Care Res. 2013, 65, 1643–1665. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Meyer, C.A.; Wesseling, M.; Corten, K.; Nieuwenhuys, A.; Monari, D.; Simon, J.P.; Jonkers, I.; Desloovere, K. Hip movement pathomechanics of patients with hip osteoarthritis aim at reducing hip joint loading on the osteoarthritic side. Gait Posture 2018, 59, 11–17. [Google Scholar] [CrossRef] [PubMed]
- Hu, R.; van Velthoven, M.H.; Meinert, E. Perspectives of People Who Are Overweight and Obese on Using Wearable Technology for Weight Management: Systematic Review. JMIR mHealth uHealth 2020, 8, e12651. [Google Scholar] [CrossRef] [PubMed]
- Gagnon, M.P.; Ngangue, P.; Payne-Gagnon, J.; Desmartis, M. m-Health adoption by healthcare professionals: A systematic review. J. Am. Med. Inform. Assoc. 2016, 23, 212–220. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Hunt, M.A.; Charlton, J.M.; Esculier, J.F. Osteoarthritis year in review 2019: Mechanics. Osteoarthr. Cartil. 2020, 28, 267–274. [Google Scholar] [CrossRef]
- Johnson, W.R.; Alderson, J.; Lloyd, D.; Mian, A. Predicting athlete ground reaction forces and moments from spatio-temporal driven CNN models. IEEE Trans. Biomed. Eng. 2018, 66, 689–694. [Google Scholar] [CrossRef] [Green Version]
- Lim, H.; Kim, B.; Park, S. Prediction of lower limb kinetics and kinematics during walking by a single IMU on the lower back using machine learning. Sensors 2020, 20, 130. [Google Scholar] [CrossRef] [Green Version]
- Tanghe, K.; Afschrift, M.; Jonkers, I.; De Groote, F.; De Schutter, J.; Aertbeliën, E. A probabilistic method to estimate gait kinetics in the absence of ground reaction force measurements. J. Biomech. 2019, 96, 109327. [Google Scholar] [CrossRef]
- Stetter, B.J.; Ringhof, S.; Krafft, F.C.; Sell, S.; Stein, T. Estimation of knee joint forces in sport movements using wearable sensors and machine learning. Sensors 2019, 19, 3690. [Google Scholar] [CrossRef] [Green Version]
- De Brabandere, A.; Emmerzaal, J.; Timmermans, A.; Jonkers, I.; Vanwanseele, B.; Davis, J. A Machine Learning Approach to Estimate Hip and Knee Joint Loading Using a Mobile Phone-Embedded IMU. Front. Bioeng. Biotechnol. 2020, 8, 320. [Google Scholar] [CrossRef] [Green Version]
Activity | Duration (s) |
---|---|
Walking | 5608 |
Ascending stairs | 564 |
Descending stairs | 420 |
Sit-to-stand | 282 |
Stand-to-sit | 283 |
Jogging | 4344 |
Cycling | 3096 |
Prediction: | Walking | Asc. Stairs | Desc. Stairs | Sit-to-Stand | Stand-to-Sit | Jogging | Cycling |
---|---|---|---|---|---|---|---|
Ground Truth: | |||||||
Walking | 0.9618 | 0.0194 | 0.0134 | 0.0000 | 0.0000 | 0.0000 | 0.0054 |
Asc. stairs | 0.2688 | 0.6452 | 0.0430 | 0.0000 | 0.0000 | 0.0027 | 0.0403 |
Desc. stairs | 0.1161 | 0.1000 | 0.7258 | 0.0000 | 0.0000 | 0.0548 | 0.0032 |
Sit-to-stand | 0.0000 | 0.0035 | 0.0071 | 0.8901 | 0.0922 | 0.0000 | 0.0071 |
Stand-to-sit | 0.0035 | 0.0035 | 0.0000 | 0.0813 | 0.8975 | 0.0000 | 0.0141 |
Jogging | 0.0266 | 0.0000 | 0.0002 | 0.0002 | 0.0002 | 0.9727 | 0.0000 |
Cycling | 0.0271 | 0.0199 | 0.0010 | 0.0003 | 0.0020 | 0.0000 | 0.9498 |
Joint Loading Impulse s | Joint Loading Profile (pt) | ||||||
---|---|---|---|---|---|---|---|
Controls | HipOA | KneeOA | no-OA | HipOA | KneeOA | ||
Walking | Hip | 192.77 | 175.74 | 183.14 | 1 | 0.91 | 0.95 |
Knee | 140.83 | 132.36 | 140.92 | 1 | 0.94 | 1 | |
Ascending stairs | Hip | 213.34 | 233.28 | 274.51 | 1.11 | 1.21 | 1.42 |
Knee | 229.2 | 235.43 | 242.27 | 1.63 | 1.67 | 1.72 | |
Descending stairs | Hip | 215.26 | 223.51 | 290.58 | 1.12 | 1.16 | 1.51 |
Knee | 218.18 | 225.02 | 240.63 | 1.55 | 1.6 | 1.71 | |
Sit Down | Hip | 136.34 | 133.68 | 154.51 | 0.71 | 0.69 | 0.8 |
Knee | 248.87 | 240.45 | 234.37 | 1.77 | 1.71 | 1.66 | |
Stand Up | Hip | 124.95 | 111.03 | 129.74 | 0.65 | 0.58 | 0.67 |
Knee | 221.54 | 199.7 | 199.74 | 1.57 | 1.42 | 1.42 |
Low-Fidelity Prototype Testing | ||||
---|---|---|---|---|
Smartphone App | Clinical Dashboard | |||
Prototype I | Prototype II | Prototype I | Prototype II | |
N | 5 | 5 | 9 | 10 |
Physical therapist/surgeon | - | - | 5/4 | 5/5 |
Age (range) | 44–68 | 49–79 | 25–60 | 27–36 |
Sex (F/M) | 3/2 | 2/3 | 4/5 | 2/8 |
Technology use (range) | 5–9 | 5–10 | 7–9 | 7–10 |
High-Fidelity Prototype Testing | |||
---|---|---|---|
Smartphone App | Clinical Dashboard | ||
Prototype I | Prototype II | Prototype I | |
N | 5 | 5 | 5 |
Physical therapist/Surgeon | - | - | 5/0 |
Age (range) | 54–68 | 52–61 | 27–44 |
Sex (F/M) | 2/3 | 2/3 | 2/3 |
Technology use (range) | 6–8 | 4–10 | 6–10 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Emmerzaal, J.; De Brabandere, A.; Vanrompay, Y.; Vranken, J.; Storms, V.; De Baets, L.; Corten, K.; Davis, J.; Jonkers, I.; Vanwanseele, B.; et al. Towards the Monitoring of Functional Status in a Free-Living Environment for People with Hip or Knee Osteoarthritis: Design and Evaluation of the JOLO Blended Care App. Sensors 2020, 20, 6967. https://doi.org/10.3390/s20236967
Emmerzaal J, De Brabandere A, Vanrompay Y, Vranken J, Storms V, De Baets L, Corten K, Davis J, Jonkers I, Vanwanseele B, et al. Towards the Monitoring of Functional Status in a Free-Living Environment for People with Hip or Knee Osteoarthritis: Design and Evaluation of the JOLO Blended Care App. Sensors. 2020; 20(23):6967. https://doi.org/10.3390/s20236967
Chicago/Turabian StyleEmmerzaal, Jill, Arne De Brabandere, Yves Vanrompay, Julie Vranken, Valerie Storms, Liesbet De Baets, Kristoff Corten, Jesse Davis, Ilse Jonkers, Benedicte Vanwanseele, and et al. 2020. "Towards the Monitoring of Functional Status in a Free-Living Environment for People with Hip or Knee Osteoarthritis: Design and Evaluation of the JOLO Blended Care App" Sensors 20, no. 23: 6967. https://doi.org/10.3390/s20236967
APA StyleEmmerzaal, J., De Brabandere, A., Vanrompay, Y., Vranken, J., Storms, V., De Baets, L., Corten, K., Davis, J., Jonkers, I., Vanwanseele, B., & Timmermans, A. (2020). Towards the Monitoring of Functional Status in a Free-Living Environment for People with Hip or Knee Osteoarthritis: Design and Evaluation of the JOLO Blended Care App. Sensors, 20(23), 6967. https://doi.org/10.3390/s20236967