Symmyo

Inspiration

We were inspired by the countless patients who finish physical therapy still struggling with weakness, imbalance, and limited motion. From stroke survivors to car accident victims, many never reach their true potential because their rehabilitation ends too early, or because their exercises are based on subjective observation rather than measurable muscle data. As engineers, we asked:

What if we could give patients and therapists clear, objective insight into how muscles actually recover — in real time?

What it does

Symmyo is an intelligent rehabilitation system that detects muscle asymmetry through surface electromyography (EMG) and recommends personalized exercises powered by our proprietary algorithm.

How we built it

Our prototype pipeline processes EMG signals through:

Noise filtering and rectification: We applied bandpass filtering and absolute value transformation to clean the signal.

Asymmetry computation: To detect imbalance between left and right muscles, we used a threshold-based ratio:

𝐴 = (𝑅−𝐿)/(𝑅+𝐿+𝜀) ​ Exercise mapping: Using a predefined database linking muscles to movements, the algorithm recommends the most effective corrective exercise.

The user interface displays asymmetry visually and suggests targeted stretches or strength exercises — creating a personalized recovery loop.

Challenges we ran into

Turning a clinical problem into a workable solution was harder than expected. We had to understand not only how muscles behave after injury.

Accomplishments that we're proud of

We’re proud that we transformed a broad healthcare problem into a cohesive, evidence-based solution that resonates with both clinicians and patients.

What we learned

We learned how fragmented current rehabilitation care can be — often limited by time, cost, and subjective judgment. The process showed us the value of objectivity and feedback in recovery, and how even small insights into muscle performance could prevent long-term complications.

What's next for Symmyo

We plan to move from concept to early validation by:

Conducting interviews with therapists, clinicians, and patients to refine real-world needs

Designing a low-cost EMG testing setup for proof-of-concept signal collection

Building a prototype of the software pipeline for detecting muscle imbalance

Seeking mentorship and accelerator support through UCSD’s innovation network

Our goal remains simple but ambitious: make rehabilitation more personal, measurable, and motivating — so that patients don’t just survive injury, but recover as much as possible to enjoy life :D.

Built With

• emg
• python