Week 20 (03/28/2021)

From March 31, 2021 to April 02, 2021, is Ceaser Chavez Day followed by Easter Break for Loyola Marymount University which means that campus will not be open to the team. The team attempted to connect multiple electrodes to Tyler's scalp to collect data, but when the data being collected appeared to provide data that appeared erroneous. Through technical analysis, the team concluded that the multiplexer switching appears to have provided extremely large irregularities in the data. At the end of the day, in order to make the most of Easter Break, data was collected from Tyler answering yes questions following by a recording of him answering no questions while the electrodes were attached to the Fp1 and Fp2 locations of the 10-20 system (frontal lobe). Data was also collected from Mark answering three minutes worth of yes questions, three minutes of no questions, three minutes of watching funny videos, and three minutes of watching sad videos. This process was done over three times. In total, over 55 minutes worth of data was collected on Tuesday.

Tyler performed more data collection on himself to collect more data on him answering yes questions, no questions, watching funny videos, and watching sad videos, making so that the team has collected over an hour worth of data for analysis. Figure 1 illustrates most of the text files of the collected data.

Figure 1: Text files over the collected data from frontal lobes

The team performed analysis on the data, primarily the yes and no files, to comprehend how the data from mentally answering yes questions differs from the data from mentally answering no questions. Figure 2 illustrates one of the analyses that was performed. The blue orbs represent the "yes" data combined, while the orange orbs comprise of the "no" data combined. The analysis method in Figure 2 illustrates the data in the time amplitude and FFT amplitude domain because the frequency domain for all the points, for some strange reason which require further analysis, is 0. After further analysis and multiple debugging methods, another method of analyzing the data was done to analyze the data in the frequency and FFT amplitude domain. Figure 3 illustrates this other method.

Figure 2: Method of analysis with 0 frequency

Figure 3: Method of analysis that provides varying frequency

THIS WEEK

Tyler

• Attempted to configure the multiplexer with all of the EEG electrodes

• Reviewed how to correct multiplexer data irregularities

• Collected data on himself mentally answering yes/no questions and from watching funny/sad videos

Mark

• Made edits and revisions to the MATLAB code for signal analysis, discrete wavelet transform, machine learning k-means, and data clustering

• Collected data on himself mentally answering yes/no questions and from watching funny/sad videos

NEXT WEEK

Tyler

• Review how to connect all the electrodes without having data irregularities from the multiplexer switching

• Collect further data using the EEG helmet for further analysis

Mark

• Perform k-means mapping to get another form of analysis of the collected data

• Collect further data using the EEG helmet for further analysis

Memo: Dr. Hossein Asghari assisted Tyler and Mark to determine the error the was being received by the multiplexer. We connected the rest of the unused channels to ground to see if that was the issue, unfortunately, that did not resolve the issue.