NeuroSymphony

• 

  NeuroSumphony

NeuroSymphony is a python application that uses neurological feedback from the Muse 2 to help a user go into or maintain a meditative/focused state.

We first transform the brain activity obtained by Muse2 from the time domain to the frequency domain using Fast Fourier Transforms.

We perform power band binning and filtering to clean data from possible noise and powerline interference and to classify the different powerbands (alpha, beta, delta, gamma).

We calculate the alpha/beta brainwave ratio obtained from the power band binning to determine the calmness of the user and adjust the volume of different soundscapes sounds to keep the user in a focused/meditative state.

The PyQt5 library is used for the GUI, the BrainFlow and numpy libraries for filtering and processing the data, the seaborn and Matplotlib libraries to plot the data, the selenium web driver to adjust the volume sliders of the soundscapes.

The program can be started from a menu screen. A graph of the real time brain wave activity along with the alpha/beta ratio plots obtained from the Muse2 sensors will be displayed in a new window.

By conducting experiments, we found that the threshold between the calm and focused state is an alpha/beta ratio of 0.7, which was determined using a 90% confidence interval.

We created an intelligent system (AI) based on a reinforcement learning algorithm called Q-learning that uses an Epsilon-Greedy policy. These algorthims optimize the system soundscapes to help the user be in the desired focused state of mind.

Built With

• machine-learning
• numpy
• python
• qt