The motivation of this clip is to demonstrate that non-electrical bio-signals can be superior to the electrical ones. Imagine you want to measure the muscle tremor of the hand. You can use EMG or directly measure its movement which is shown here.
EMG measures muscle activity which corresponds roughly to force. From physics we know what F=ma meaning that force is proportional to acceleration. An alternative to EMG is thus acceleration which is presented here. The accelerometer used is an ADXL103CE by analog devices which is one of the most sensitive ones on the market but requires only two external components (one blocking capacitor and a 1st order lowpass). While the EMG is slightly noisy the accelerometer provides a virtually noise-free signal allowing to measure very small finger movements and even muscle tremor.
Note that there is some crosstalk between the two channels because the accelerometer impacts slightly on the EMG baseline which shows up as a DC drift. However, the EMG itself has much higher frequencies which can clearly be seen when Vasso is squeezing the tennis ball. This shows again that it is important to interpret the signals properly to be able to distinguish between crosstalk, noise and signal.