FP-00019

Section 1 - Basic information about you and your application:

Title of research project
Can smartphone sensors be used to assess and train residual muscle strength deficits after knee injuries?

Grant Type
The ORUK Early-career Research Fellowship

Research area
Diagnostic and treatment

Duration
24

Start date
April 1, 2024

Have you previously received funding from ORUK?
No.

Profession
Academic scientist

Your current job title/position
Assistant Professor in Musculoskeletal Pain and Rehabilitation

Are you an early-career researcher (ECR)? (definition of ECR)
yes

Section 2 - Lay summary

Lay summary:

Approximately 64 people out of 100,000 experience an Anterior Cruciate Ligament (ACL) injury in their lifetime. In the UK, ACL reconstructions have increased 10-fold between 1997 and 2017, reaching an incidence of 24.2 every 100,000 people. These injuries are potentially career-ending for athletes since they require longer than a year to return to sport and to performance, and may result in knee osteoarthritis in the long term.

To assess an individual’s risk of injury and whether they are ready to resume their daily activities or to return to sport, clinicians assess how they move. A key issue is that movement is often assessed visually, or using laboratory equipment that is expensive and usually not available outside specialized centers. In addition, home exercises programs are often prescribed for prevention or rehabilitation. However, it is difficult for clinicians to monitor their patient’s progress and decide if the exercises are effective. Inexpensive, easy-to-use, remote solutions to assess how people move would help clinicians gather objective information to guide their clinical decisions.

In this application, we propose to use smartphone sensors to assess and monitor remotely how people move. We have previously demonstrated that smartphone sensors are almost as accurate as laboratory equipment when assessing how people perform a landing stabilization task, an exercise widely used for prevention and rehabilitation that consists in jumping, landing, and regaining balance as quickly as possible.

We will recruit people who have recovered from knee injuries, and ask them to use their own smartphone to test themselves at home. We will determine whether their jumping performance is comparable between the injured and non-injured leg. We expect that the smartphone sensors will measure worse performance for the previously injured leg.

We will also recruit people with marked differences in jump height between legs. Participants will perform landing stabilization exercises on their weaker leg for 6 weeks, recording every session using their smartphone. We will use this data to understand when the exercise is not effective anymore (for instance: jump height stops improving). Clinicians will be able to use this data to provide individualized exercise programs.

Participants will provide feedback on the app used to collect data and their experience with the overall procedure, so that we can identify any barriers to the implementation of the proposed procedure.

This Fellowship will provide clinicians with a free, widely-available tool to guide their clinical decisions by assessing movement objectively and remotely. This will result in lower incidence of knee injuries and improved rehabilitation outcomes in the long term. Demonstrating the effectiveness of this approach will be the stepping stone for the use of smartphone sensors in other areas of musculoskeletal health, for prevention, diagnostics and intervention.

Section 3 - Purpose of research

Purpose of research:

The aim of this research project is to determine whether smartphone sensors can provide objective motion data to support diagnosis and treatment of individuals who sustained knee injuries.

The first objective (Work package 1) is to determine whether smartphone sensors can be used to identify residual motor deficits in individuals who have been deemed recovered from knee injuries. By asking participants to perform a landing stabilization test while measuring themselves using their own smartphone, we expect that primary (jump height) and secondary outcomes (impact forces at landing, postural sways, trunk orientation) will be worse for the leg that sustained an injury compared to the leg that did not.

The second objective (Work package 2) is to determine whether smartphone sensors can be used to monitor home exercise programs in individuals who sustained a knee injury (for return to performance after rehabilitation) and in healthy people (for injury prevention). Participants with or without previous knee injuries who demonstrate side differences in jump height >10% will perform a jump-focused exercise program at home, monitoring every session using their own smartphone. Their feedback on the usability and barriers of the currently available app will be collected using a questionnaire. By analysing the session-by-session improvement of jump height, we will establish: 1) adherence to the intervention; 2) when the improvement plateaus for different people; 3) whether adherence and performance determine the overall exercise performance improvement.

These work packages will demonstrate that smartphone sensors can be used as a free tool to: 1) identify residual motor deficits remotely; 2) identify non-responders at an early stage of the exercise program to provide individualized exercise prescription. This information will be shared in open-access publications and conference presentations (at least one for each work package), supplemented by educational material and other dissemination activities.

Section 4 - Background to investigation

Background to investigation:

LITERATURE REVIEW:

Health implication of sport injuries. Sport injuries influence the overall health of the community. Injuries are one of the main reasons for dropping out of sports at a young age, contributing to the limited amount of physical activity observed in adolescents [1]. Sport injuries also increase the likelihood to develop osteoarthritis [2], pathology that costed the NHS more than £100 billion between 2018 and 2028 [3]. To promote sport participation and physical activity, and to reduce prevalence of joint degeneration in the long term, professionals and clinicians should aim to reduce the incidence of sports injuries and resolve movement deficits that persist after injury. In this application, we propose to investigate whether smartphone sensors can help clinicians identify residual movement deficits and monitor exercise performance remotely, both for injury prevention and return to sport.

Return to sport. Strength and movement deficits can be observed long after an athlete returns to sport after injury [4]. However, these deficits are not always detected by the recommended clinical tests. Recent evidence shows that athletes cleared to return to sport because of symmetrical performance of the hop distance test, a test recommended to determine whether an athlete is ready to return to sport after Anterior Cruciate Ligament (ACL) reconstruction [5], still demonstrate a 20% deficit of jump height comparing the injured and non-injured leg because of residual muscle strength deficit. Objective assessment of jumping performance may therefore guide exercise selection in the rehabilitation process and help define clearer return to sport criteria.

Exercise. There is strong evidence that injury prevention programs that include strength and balance training reduce the incidence of sport injuries [6]. However, systematic reviews [7,8] demonstrate that exercise works only if performed (quantity of exercise), and only if performed well (quality of exercise). First, injury prevention programs reduce the incidence of ACL lesion by 64% when adherence is high, but they have no effect when adherence is medium or low [7]; this demonstrates how measuring adherence is key to ensure that prevention programs are implemented correctly, and hence prevent injuries. Meta-analyses also demonstrate that a standardised exercise program (FIFA 11) does not reduce injuries, whereas a similar program but with exercises tailored to the athlete (FIFA 11+) reduces the incidence of lower limb injuries by 39% [8]. This highlights how monitoring exercise performance and prescribing individualized exercise programs is key for injury prevention. Taken together, this evidence suggests that both quantity and quality of the exercise are important for effective injury prevention.

Issues with current tests. Objective strength and movement tests for return to sport can only be performed in specialized centers. Home exercise adherence is often assessed via self-reported logs. However, self-reported logs underestimate sedentary behaviour by 1.74 h/day [9] and overestimate the amount of exercise performed by almost 20% [10] when compared to objective, sensor data. In this application, we propose to investigate whether smartphone sensors can be used to identify residual muscle strength deficits in individuals who sustained knee injuries (Work package 1), and to monitor quantity and quality of home exercise remotely (Work package 2). This will support the proposed approach to objectively assess residual movement alterations and to prescribe individualized exercise progression based on zero-cost, remote assessment of motor performance.

PERSONAL TRACK RECORD:

My research and teaching focuses on the effect of pain and injury on movement, and its implications for rehabilitation. Past and current research include studying motor control and strength deficits in women with patellofemoral pain, published in Medicine and Science in Sport and Exercise, and I have obtained competitive funding in the UK to investigate motor adaptation to experimental pain. To translate these laboratory findings into practice I have recently started a new research line focused on measuring movement objective using smartphone sensors. My teaching includes modules on training and motor performance, where all the practical sessions involve using smartphones to measure movement. The potential relevance of this interactive teaching framework has attracted the attention of a professional organization of clinicians abroad, who invited me to present my teaching activities in a congress and in a webinar. The collaborator included in the application will help me develop in the field of exercise physiology and sport, ensuring that the tests can be implemented in practice to maximize impact.

My goal as a scientist is to understand whether objective analysis of movement can improve how we assess and treat individuals with musculoskeletal pathologies. My vision is to achieve this by running an experimental research program in the laboratory alongside an applied research program in the clinic. This Fellowship will provide strong support for the applied side of my work, allowing me to translate the laboratory findings of my group in practice. The implementation of smartphone sensors in practice would also help clinicians working in the field of musculoskeletal sciences develop a data-oriented mindset, potentially contributing large motion analysis datasets to the NHS and other researchers.

PRELIMINARY DATA:

In preparation for this project, we investigated whether smartphone sensors provide motion estimates that are valid compared to gold-standard laboratory techniques (preliminary data). When compared to force plates, smartphone accelerometers provided valid measures of time of flight (N=52; Intraclass Correlation Coefficient: 0.97 [0.94 – 0.98]; figure attached). Secondary outcome measures such as landing impact, postural sways, and trunk orientation at landing all exceeded ICC=0.75, demonstrating good validity. The between-day reliability of jump performance metrics will be investigated in the current study, but we have already established good-excellent reliability for body orientation measured with smartphone sensors during different tasks in two different publications [references blinded]. The reliability data was collected remotely, which demonstrates feasibility for the projects proposed in the current application.

REFERENCES:

1. Emery CA. Best Pract Res Clin Rheum 2019.
2. Culvenor AG. BJSM 2014.
3. Woolf A. https://www.england.nhs.uk/blog/tackling-the-elephant-in-the-room/
4. Tayfur B. Sports Med 2021.
5. Kotsifaki A. BJSM 2022.
6. Lauersen JB. BJSM, 2014.
7. Halvorsen KC. HSSJ, 2022.
8. Thorborg K. BJSM, 2017.
9. Prince SA. Int J Behav Nutr Phys Act, 2020.
10. Nicolson PJA. JOSPT 2018

Section 5 - Plan of investigation

Plan of investigation:

Design. WP1: within-subject comparison, where outcome measures will be compared between injured and non-injured leg. WP2: longitudinal study, where changes in outcome measures will be compared before and after a 6-week home exercise program and monitored throughout the intervention.

Participants. To determine whether smartphone sensors can identify residual motor deficits, in WP1 we will recruit 50 participants who have had surgery after a knee injury in the last 10 years and have been cleared to return to normal daily living activities and/or to sport. Exclusion criteria include current injury or pain. Based on an effect size of 0.67 on side differences in jump height in athletes cleared to return to sport after ACL reconstruction [5], power of 80% and alpha of 0.05, a sample size calculation resulted in 20 participants. We increased the number to 50 to account for variability in the type of injury/surgery (not exclusive to ACL reconstruction to facilitate ecological validity). At least half of the sample size will be participants who have had an ACL reconstruction. For WP2, we will recruit participants who demonstrate asymmetrical jump height between left and right leg (e.g.: difference >10%); these participants may have recovered from a knee injury (see WP1) or be healthy participants. Preliminary data demonstrated that 8/52 healthy participants had a side difference in jump height >10%. A sample size calculation using an effect size of 0.8 based on two meta-analyses that summarized the effect of plyometric training on jump height [11,12], power of 80% and alpha of 0.05 revealed that a minimum of 15 participants are necessary. Since we plan to explore the variability in the response associated to adherence and quality of the exercise, and in people with/without previous knee injuries, we have decided to increase the sample size to at least 50 participants. Participants will be recruited by word of mouth, physical flyers, online ads and we will contact the UK National Ligament Registry to advertise the study through the registry.

Protocol. For both WPs, we will collect information on participant characteristics, activity levels (e.g.: Tegner score) and function (e.g.: Knee Injury and Osteoarthritis Outcome Score). Participants will perform single-leg landing stabilization exercises either as a test (WP1 and WP2) or as intervention (WP2). This exercise evaluates several movement characteristics (muscle strength, neuromuscular strategies, balance), it is used as a screening tool for ACL injuries in practice [13], and meta-regressions showed that ACL injury prevention programs that include these exercises are more effective in reducing the risks [14]. Participants will be asked to: ‘stand on one leg, jump as high as possible, land on the same leg, and stabilize your balance as quickly as possible’. Participants will hold their own smartphones on the chest, which we have shown to result in valid measures (see preliminary data).

For WP1, individuals will participate in a single session. After a self-selected warm-up, participants will perform the landing stabilization exercise twice on the injured and non-injured leg, randomized, 5 repetitions each. The first trial for each leg will be considered a practice trial, and the second set of 5 repetitions will be considered the test trial. Data will be collected remotely, under supervision of a researcher using Zoom.

For WP2, participants will perform the test above in two sessions a week apart to establish their baseline and the between-day reliability of the outcome measures. Then, participants will perform the landing stabilization exercise for 6 weeks, 15 repetitions, 3 times a week. Participants will monitor exercise performance in every session using their own smartphone. At the end of the intervention, the test will be performed again. The sessions before and after the intervention will be supervised by a researcher over Zoom. Participants will then complete a questionnaire to provide feedback about the use of the app.

Outcome measures. The primary measure will be lower limb muscle strength, estimated as time of flight during maximal jump height. Secondary measures will be neuromuscular control (impact landing forces), posture (trunk orientation) and balance (postural sways). These will be measured using the sensors in the participant’s own smartphones using a free app, such as Matlab Mobile or Phyphox. These apps record data from the smartphone sensors at 100 Hz. Data is either uploaded to a server (Matlab Mobile) or shared via email (Phyphox). These procedures have been approved by our IT department through a Data Protection Impact Assessment.

Analyses. In both WPs, better exercise performance will be defined as: higher time of flight (jump height), lower landing impact forces, less trunk inclination, fewer postural sways. In WP1, performance of injured and non-injured leg will be compared using paired T-tests; since individuals who recovered from ACL injuries have more long-lasting deficits in motor performance [4], further analyses will be performed in this subgroup. In WP2, adherence will be quantified as the percentage of jumps performed; exercise performance will be quantified as the change from baseline. Between-day reliability will be quantified using Intraclass Correlation Coefficient between the data collected in the first two sessions. Plateau of performance improvement will be identified as the first of three consecutive sessions where the performance (e.g.: jump height) does not exceed the 95% confidence interval of the previous session. Overall improvement will be calculated as the change in performance comparing before/after the intervention. Regression analyses will be used to determine whether lower overall performance improvements are associated to lower adherence and earlier plateau of performance improvement.

Risks for clinical adoption and commercial exploitation. This project has minimal risks. In terms of commercialization, while it is possible that ad-hoc apps may be developed based on the findings of this study, the existence of free apps (e.g.: the ones used in this study) and the possibility to analyse the data using widely available software (e.g.: MS Excel) ensures that clinicians and other practitioners will be able to access the procedure described in this study for free.

REFERENCES:

11. Ramirez-Campillo R. J Sports Sci. 2020a
12. Ramirez-Campillo R. J Sports Sci. 2020b
13. Petushek EJ. AJSM, 2015
14. Petushek EJ. AJSM, 2019

Section 6 - Research environment and resources

Research environment and resources:

Because of the applied and remote nature of the project, laboratory facilities will not be needed. The main support provided by our institution concern data security. When using the Matlab Mobile app linked to an institutional email account, the data is securely stored in a secure server, which satisfies the our data protection policies. Participants’ personal information will be securely stored on institutional servers using REDcap, whose license is paid by our institution.

For the recruitment of participants, we have close relationships with our sports department (including staff with cross-appointments) who will facilitate recruitment of healthy participants as well as those who have recovered from knee injuries. We have previously recruited 30-50 healthy participants in about 4 months in four different studies, and had to exclude several potential participants because of previous knee injuries. Given the remote nature of the project, in addition to this local recruitment, we will be able to recruit through a large network of connections with sporting clubs and other universities through the graduate research program led by the collaborator in this application. In our department we currently have several PhD students embedded in sports clubs and organization, which demonstrates that we can handle remote data collection effectively.

Our institution also offers several services to assist with the dissemination and application of our research findings, both in the academic and non-academic environment. Publication in high-quality, open-access journals will be covered through our institutional agreements. Public Engagement and Impact offices will support our dissemination and impact strategies.

Section 7: Research impact

Who will benefit from this research?

The main beneficiaries of this research will be patients and clinicians. Establishing the use of smartphone sensors as a tool to assess movement objectively will improve the health of patients by contributing to prevent injuries in healthy individuals and help determine objective return to sport criteria. The fact that the proposed technology is zero-cost and already widely available means that it can ultimately be scaled up have impact at national and worldwide levels. Since most people own a smartphone, our proposed approach also helps remove barriers related to wealth, fostering inclusivity.

Clinicians will benefit by having objective, recorded motion data that they can use to inform their clinical decisions, and to review performance before/after interventions. The remote assessment and monitoring aspect of the project means that the tests can be performed from home, further facilitating the implementation in practice and possibly contributing to decreasing waiting time in the NHS.

How can your research be translated in real-life?

The objective movement analysis proposed in this application is simple and zero-cost, therefore accessible to most clinicians and other practitioners for free and with minimal time investment. All the movement characteristics investigated in this application are currently part of the practical sessions of my module focused on motion analysis. Year 3 undergraduate students learn how to collect and analyse the data within an hour, which demonstrates the feasibility of learning how to apply the procedure in real life. Results of this project will also feed directly into my teaching material, which will be studied by at least 250 students every year. In addition to the scientific publications, which will be in open access journals, we will publish video examples of data analysis which will be useful for clinicians interested in learning how to analyse the data using a digital spreadsheet (e.g.: MS Excel).

How will your research be beneficial for ORUK and its purpose?

We believe that our proposed application aligns with several of ORUK’s priorities.

Beyond surgery. The long-term goal of this project is to contribute to reduce the amount of knee injuries and re-injuries (and therefore surgery) by improving prevention, adherence and personalized interventions through objective motion analysis.

Faster impact. Since most people own a smartphone, our proposed approach can be implemented at zero cost and with minimal time investment. This has the potential to lead to immediate impact, potentially at a world-wide level.

Physiotherapy. We believe that the proposed project can contribute to make physiotherapy more data-oriented, facilitating communication with medical specialties where data culture is more developed. This is in addition to assisting professionals who work in injury prevention and return to sport.

Focus on diagnosis and treatment. The proposed project focuses on both diagnosis (identification of motor deficits) and treatment (remote exercise monitoring), both pointing towards precision medicine.

Section 8: Outreach and engagement

Educational material. Open-access scientific publications will be complemented by educational materials such as video and tutorials on how to collect and analyse the data using MS Excel or similar. This will assist clinicians and other practitioners willing to implement the procedures described.
Teaching. The results of the project will be included in our teaching activities. This means that every year 250+ students in different programs (undergraduate and master in physiotherapy and sport science) will learn about how to use this technology and could implement it in their future careers, having a huge potential impact. The open access publications and educational material will ensure that other academics and teaching staff will have the possibility to include the results of this study in their teaching, therefore increasing knowledge and skills even further.
Network. Results will also be widely presented at formal and informal meetings organized through our network. In the last two years, I have been invited to speak about using smartphones to measure movement in practice at two events organized by a physiotherapy association abroad. Similar opportunities will be sought through local connections, especially through the connections of the graduate sport program led by the collaborator of this application and our sport department.
Public and Business Engagement. To maximize the reach of this research, we will liaise with our Public Engagement and Communication offices to create an outreach plan. This will ensure that our research is impactful while establishing myself as a leader in the field.

Section 9: Research budget

Requested funding from ORUK

University fees (if any)
£0

Salary
£97904.61

Consumables
£5000

Publications
£0

Conference attendance
£3000

Other items
£2500

Total 'requested fund'
£108404.61



Other items
ADDITIONAL ITEMS: OTHER DIRECTLY INCURRED: £1,500. To cover PDRA recruitment costs, including advertising and travel. OTHER DIRECTLY INCURRED: £1,000. To cover study advertisement fees and other generic costs. JUSTIFICATION OF RESOURCES FOR THE AMOUNTS IN 'REQUESTED FUNDING FROM ORUK' (ABOVE): SALARY: £97,904.61. I request a Grade 7.29 PDRA for 22 months to carry out the data collection and analysis for Work Packages 1 and 2. The PI’s salary will be covered by the institution. CONSUMABLES: £5,000. To compensate individuals to participate in the experimental studies of Work Packages 1 and 2. PUBLICATIONS: £0. Covered by institutional agreements. CONFERENCE ATTENDANCE: £3,000. To cover conference fees, transports, accommodation, meals for at least two conferences (e.g.: International Society of Biomechanics, European College of Sport Sciences, Physio UK).



Other secured funds

Internal funding
£0

Partner (University)
£0

Partner (Commercial)
£0

Partner (Charity)
£0

Other sources
£0

Total 'other funds)
£0

Section 10: Intellectual property and testing on animal

Is there an IP linked to this research?
No

Who owns and maintains this patent?


Does your research include procedures to be carried out on animals in the UK under the Animals (Scientific Procedures) Act?
No

If yes, have the following necessary approvals been given by:

The Home office(in relation to personal, project and establishment licences)?


Animal Welfare and Ethical Review Body?


Does your research involve the use of animals or animal tissue outside the UK?
No

Does the proposed research involve a protected species? (If yes, state which)



Does the proposed research involve genetically modified animals?


Include details of sample size calculations and statistical advice sought. Please use the ARRIVE guidelines when designing and describing your experiments.


There should be sufficient information to allow for a robust review of any applications involving animals. Further guidance is available from the National Centre for the Replacement, Refinement and Reduction of Animals in Research (NC3Rs), including an online experimental design assistant to guide researchers through the design of animal experiments.

Please provide details of any moderate or severe procedures


Why is animal use necessary, are there any other possible approaches?


Why is the species/model to be used the most appropriate?