Project
| # | Title | Team Members | TA | Documents | Sponsor |
|---|---|---|---|---|---|
| 44 | Reusable Muscle Activation and Degenerative Disease IoT Sensor |
Branden Youssef Caleb McEwen |
Kexin Hui | design_document0.pdf design_document0.pdf design_document0.pdf design_document0.pdf final_paper0.pdf presentation0.pptx proposal0.pdf |
|
| Caleb McEwen - cemcewe2 Branden Youssef - byousse2 Idea Post - https://courses.engr.illinois.edu/ece445/pace/view-topic.asp?id=26994 - Problem - There are numerous protocols suggested for lifters to improve their strength and muscle size, but it can take months to learn how to perform them with proper form, or what set and rep schemes are ideal for the person. Skills like the mind-muscle connection and form have to be learned by trial and error, but could be learned much faster using EMG data. Advanced lifters could determine if reducing the weight and slowing down the reps on their exercises produce equivalent or improved muscle activation, and could then confidently perform those slower reps, which provide less of an injury risk. Additionally, people who are developing degenerative muscle diseases may not know about it for months after the onset of a disease. Earlier detection with EMG sensors allows for improved prognoses and reduced medical bill and insurance costs. - Solution - Our project is an inexpensive, reusable EMG sensor that gives user's muscle activation data for optimizing their workouts themselves, on the go. It will also be able to refer user's to see a doctor if it detects fibrillations that indicate a certain class of muscular degenerative disease. It will consist of electrodes attached to an Atmel microprocessor that will send data to the user's phone for viewing in an app. The microprocessor will connect to a bluetooth modem, filters, a mixer, and the electrodes so the data received in the Atmel chip is filtered and amplified to be comprehensible for further processing. The processing system will be powered by a button cell battery. The EMG requires two electrodes to detect one muscle's activation, so the electrode that is not connected to the microcontroller (instead on a bone or joint; not on the muscle belly) will connect to the rest of the sensor using flat wires. The sensor would come with pre-cut gauze squares and 90% rubbing alcohol for cleaning the sites of electrode placement, as well as a manual with pictures on placement for each muscle. Data visualization and a user interface will reside in the Android app. The microprocessor and hardware filters will be used to make the raw data readable for the Android phone app. The app will take readable data and separate it into differential muscle activations, as well as display it on a graph and detect fibrillations over time. An additional idea would be to include an LED that lights up on the main electrode (housing the microprocessor and lying on the muscle belly) whenever the user is achieving a set goal muscle activation value. Also, I have reached out to a professor at the U of Pittsburgh about his patent on a dry electrode and if it is on sale. If not, we will include an electrolyte solution for the user to apply to their skin after the rubbing alcohol. Sensor/Processing Subsystem - UPDATED Reusable EMG electrodes to detect muscle activation of surface muscle bellies Atmel Microprocessor Since we are dealing with a small frequency range we'll use a Low Pass Filter then use a Mixer to spread the information, then do a Band Pass Filter to allow information from the electrodes be distributed evenly for an A/D converter Amplifier (Stretch Goal) LED housed on the main electrode and connected to the Bluetooth Modem Network Subsystem - Bluetooth Modem Power Subsystem - Replaceable button cell battery connector - Criterion for Sucess - Our solution should be able to detect electrical potentials roughly between 50uV and 30mV at a rate of 7-20Hz, sending it to the user's phone at 10 samples per second (subject to change). Fibrillations come in at amplitudes of 20uV to 300uV at rates of 2 to 20Hz, so not all fibrillations will be detected, but depending on the frequency with which we can poll the Network Subsystem, we hope to detect a significant percentage of fibrillations. The system should last a few weeks on one battery, assuming 2 hours of use 5 days a week. Athos currently has clothing with EMGs included, but our system will be much cheaper and will not require the user to wear any additional clothes. |
|||||