*3.3. Signal Visualization*

Performing experiments with EEG can be very complex due to the difficulty of creating high-quality recordings. One of the main issues, apart from the design of the experiment itself, is measuring how well the data were captured. Electrodes inside devices are very sensitive to electromagnetic noise, so ensuring good skin contact is critical to avoid inconsistent results across recordings. For this reason, the library includes the necessary features to enable researchers to watch the signals of several devices in real time.

Sometimes, experiments are not run by medical experts, which is especially the case for low-cost devices. Therefore, we ensured that anyone can have access to the data while participants are being evaluated. The web server is started from the machine to which the devices are connected and returns to the console the internal IP address, together with the associated port. The process was simplified as two methods:

#### start\_streaming(search\_streams(), debug=True)

There are two different ways of visualizing the signals, compressed and expanded. In Figure 3, the overall compressed view of the website is shown with two devices at the same time. The latter is shown in Figure 4. Additionally, there are controls for the update interval (from 200 ms to 5 s), the channels to watch, the zoom level, and the possibility to play and pause the live view. For anyone without deep knowledge about performing experiments, we included a marker to know if a particular electrode had good skin contact. This helps to keep noise sufficiently low to retain a high probability of success. Lastly, the library automatically detects how many Muses are connected to the computer and adapts the interface to show those.

In Figure 5, we show our solution for experiments with multiple devices connected at the same time, while watching the streams. Additionally, the data file structure exported using the BIDS format is presented.

**Figure 5.** Solution designed for experiments.
