SmartSleeve: Multimodal Physiological Diagnostic Device for Patellar Injury Prevention

The Problem: The Subjective Gap (Figures 1-2)

Approximately 25% adults experience knee pain, and in Canada, 700 000 youth hurt their knees playing sports, and half of them end up developing osteoarthritis by the age of 40. Despite the prevalence of knee injuries, most of us rely on our own feelings to gauge how our tendons are doing. But this only leaves us vulnerable to injury.

Another barrier is the expense of accurate diagnostic tools. Currently, MRI is the most accurate diagnostic imaging modality, but they range from $600 to $1 000 along with significant wait times. MRI’s are simply inaccessible for daily monitoring. There is a clear need for a low-cost diagnostic tool that provides real-time analysis. I want to shift how we deal with knee pain from treatment-focused to prevention-focused.

Motivation:

Like many other athletes, I have experienced chronic knee pain since 2019. I felt that sometimes I pushed my knee too hard when it “felt fine”, only to suffer a setback in recovery after playing. I wanted to bridge the gap between how an injury feels and its actual internal functioning.

The Solution: The SmartSleeve

These problems are why I wanted to create the SmartSleeve. This detection device incorporates surface electromyography to measure muscle inhibition and thermography to measure inflammation. These two biological factors are synthesized to create a "readiness score" for the person.

Goal:

Create a diagnostic device for anyone who wants to know if their knees are at risk for injury.

System Design:

The SmartSleeve went through two prototypes. The first used a b.board and a microbit V2 to house the two sensors: a Grove sEMG sensor and an IR-Thermo 3. I use block-based code as a proof of concept for the diagnostic device. With this original device, I was able to create a readiness score that took into account the neuromuscular inhibition and tendon inflammation.

I transitioned to Arduino, using a TMP117 temperature sensor for higher precision and an LCD for the display. This second prototype had an interactive interface along with straps to secure the device on the patellar tendon.

How it works:

The Grove EMG detects neuromuscular inhibition in your VMO, while the TMP117 detects inflammation in your patellar tendon. Together, they create a readiness score

Testing:

Participant Pre-Assessment: 12 participants rated their knee pain on a scale of 1-10 during a single-leg squat. Below 3/10 = healthy leg. Above 3/10 or history of tendinopathy = injured leg.

Preparation: Skin was cleaned with isopropyl alcohol to reduce signal noise. Two electrodes were placed 2cm apart on the VMO, with a third on the kneecap as a ground.

Fatigue Test: 15 Bulgarian Split Squats per leg, metronome-controlled at 3 seconds per rep. Split squats were chosen to prevent compensatory loading, where the brain shifts weight to the healthy leg during standard squats.

Baseline vs Fatigue: Immediately after, peak temperature and a final MVIC were recorded on both legs to detect fatigue-induced inhibition and thermal inflammation.

Criteria for Success:

Accuracy (Correlation): Whether the SmartSleeve can match high subjective pain with low “readiness” and vice versa

Prediction (The Subjective Gap): Whether the SmartSleeve can detect physiological markers (low EMG activation and high thermal delta) in those who report low subjective pain.

Results

Correlation: There is a clear negative correlation between subjective pain rating and the device’s readiness score. Generally, as the stated pain increased, the readiness score displayed on the device decreased. There was a decrease in neural muscle activation and thermal inflammation detected in a few tests. The trendline on Graph A further illustrates this relationship, which had an R2 value of 0.503. This confirms that the device is able to accurately mirror the perceived state of the tendon using these two physiological markers.

Prediction:

However, there is still a critical difference between pain level and readiness score. This discrepancy shows the “Subjective Gap” that people with knee pain often experience. This is where the device’s sEMG detected muscle inhibition and the thermal sensor detected hyperemia, but had not impacted the participant's own pain perception. This subjective gap is seen most distinctly in participant 3. They experienced a 48.7% symmetry gap between their right and left knees after exercise, with a degree increase in the injured knee. This marked a severe functional imbalance. They still reported a pain level of 6, but often athletes will push themselves past this threshold.

In most of the participants, their injured leg slightly experienced more inhibition than their healthier counterparts. The difference in inhibition was not significant; however, when looking at participants with previous ACL tears, there was a clear muscle inhibition imbalance between the limbs.

Limitation of Results:

A key limitation of my results is the moderate R value of 0.503. This means that the pain score only explains half of the readiness score, but this is not a flaw. If my diagnostic device only mirrored the subjective pain, it would not be useful. The distinction is in finding those subjective gaps.

While this device does successfully analyze the condition of the patellar tendon using two physiological indicators, it does not give the complete story on the tendon condition. If someone experienced high pain, but the device read as a high readiness score, what would that mean? This is where the gap between internal indicators and actual physical movement indicators comes in. A person’s knee pain could be due to jumping mechanics, where their knee caves in, putting additional pressure on the patellar tendon. However, they could experience pain before any neural inhibition could be detected or inflammation could reach a significant level.

Conclusion:

The development and testing of this multimodal diagnostic tool successfully showed that a low-cost device can accurately detect the markers of knee pain. By synthesizing surface electromyography and thermography, the device was able to mirror the pain scale of participants while even giving more accurate detection than their own perception. The biggest discovery of this project was the “subjective gap” that I was able to find. This is a window where there is high injury risk, where the thermal and neuromuscular state leaves the patellar tendon vulnerable before conscious pain. The findings confirmed that NMI and hyperemia are detectable markers of patellar pain.

Additionally, the cost of this SmartSleeve was about $140 dollars, which, compared to the hundreds of dollars and wait times for MRIs, this device works as a quick check for your knee.

SmartSleeve is a usable diagnostic system for helping see beyond the subjective pain scale!

Future Work:

Integrating an MPU6050 (IMU) to correlate muscle inhibition with valgus collapse (knee caving). This would provide a 3D map of injury risk

Testing how well the SmartSleeve can detect knee injury risk in the elderly with osteoarthritis

Testing how the SmartSleeve can assess other tendons for injury risk, such as your Achilles tendon or rotary cuff tendon (shoulder)

I'd like to thank Louisie Haycocks, a member of Brilliants Labs, for supplying me with the b.board and microbit for my first prototype, as well as the sEMG click.

And thank you to all my participants for letting me test their knees!

References:

Aagaard P, Simonsen EB, Andersen JL, Magnusson SP, Halkjaer-Kristensen J, Dyhre-Poulsen P. Neural inhibition during maximal eccentric and concentric quadriceps contraction: effects of resistance training. J

Appl Physiol (1985). 2000 Dec;89(6):2249-57. doi: 10.1152/jappl.2000.89.6.2249. PMID: 11090575.

Barber-Westin, S., & Noyes, F. R. (2020). One in 5 athletes sustain reinjury upon return to high-risk sports after ACL reconstruction: A systematic review in 1239 athletes younger than 20 years. Sportsmetrics / Noyes Knee Institute. https://sportsmetrics.org/one-in-5-athletes-sustain-reinjury-upon-return-to-high-risk-sports-after-acl-reconstruction-a-systematic-review-in-1239-athletes-younger-than-20-years/

Bertrand Sonnery-Cottet, Thomas Ripoll, Etienne Cavaignac, Prevention of knee stiffness following ligament reconstruction: Understanding the role of Arthrogenic Muscle Inhibition (AMI), Orthopaedics &

Traumatology: Surgery & Research, Volume 110, Issue 1, Supplement, 2024, 103784, ISSN 1877-0568, https://doi.org/10.1016/j.otsr.2023.103784. (https://www.sciencedirect.com/science/article/pii/S1877056823003365)

Fokam D, Lehmann C. Clinical assessment of arthritic knee pain by infrared thermography. J Basic Clin

Nguyen US, Zhang Y, Zhu Y, Niu J, Zhang B, Felson DT. Increasing prevalence of knee pain and symptomatic knee osteoarthritis: survey and cohort data. Ann Intern Med. 2011 Dec 6;155(11):725-32. doi: 10.7326/0003-4819-155-11-201112060-00004. PMID: 22147711; PMCID: PMC3408027.

Physiol Pharmacol. 2018 Oct 31;30(3). doi: 10.1515/jbcpp-2017-0218. PMID: 30375348.

“Patellar Tendinitis.” Mayo Clinic, Mayo Foundation for Medical Education and Research, www.mayoclinic.org/diseases-conditions/patellar-tendinitis/symptoms-causes/syc-20376113#.