validation of strive tech sense3 muscle monitoring
In partnership with the Human Performance Research Lab at California State Polytechnic Institute, Strive validated the muscle (sEMG) functionality of the Sense3 product. Overall, the correlation between the medical grade device and Strive is 0.91, indicating a very high correlation. Standard deviation of correlations is 0.046.
The setup consisted of active and healthy subjects who performed three squats with different weights (30%, 60%, and 80% of max). The subjects were required to test while wearing both Strive Sense3 and Delsys Trigno EMG sensor. The setup required subjects to perform the squats while monitoring Quad muscles groups with sensors overlaying Rectus Femoris on both the left and right leg.
Correlations between Strive and Delsys indicates repeatable and reliable readings. Strive showed average correlation between its measurement and Delsys system across all subjects. Furthermore, the normalized data between subjects indicate correlation while having different total values. This is due to the different body compositions between subjects. Strive recommends that muscle data is normalized individually across subjects to account for different body compositions.
Average correlation by each subject
The subjects performed tasks in serial manner, repeating three reps for each set. The comparison was done by looking at the peak readings for each rep, and their comparison between Delsys and Strive. Subjects 2 and 9 showed correlation of 0.84 and 0.80 respectively and were considered the lowest correlation points. The overall correlation was done by averaging each reading for both Strive and Delsys.
Subject correlation range
Looking at the readings within each subject, subjects 2, 4, and 9 had outlier readings between 0.51 and 0.76.
Average correlation by weights
Correlations by weight vary by 0.04 range which is negligible relative to different weights.
Overall, the data showed high correlation in repeated cases between both systems. It’s important to notice that while correlation is high, 15% of samples were removed due to poor contacts between electrodes and the body. There was no significant difference in signal repetition amplitude (95%, peak, or sum) measured between systems.
The system validation was conducted over two days with all thirteen subjects.