Assignment 6 - Option C (Wearable Brain Monitor)

PSYCH 403A1 - Neuroimaging and Neurostimulation
Due: September 29, 2025
🎯
## What You'll Learn (In Plain English!)

Think of this assignment like cleaning up a noisy radio signal to hear your favorite song
better! By the end, you'll be able to:
📥 Load brain wave recordings (like opening an audio file, but for brain signals!)
🔍 Understand what EEG signals look like (spoiler: they're very squiggly!)
🔧 Apply digital "filters" (like noise-canceling headphones for brain data)
📊 Create beautiful scientific plots (that you'll be proud to show off!)
🌈 Make colorful frequency charts (showing which brain wave "notes" are
strongest)
🎵
## The Brain Wave "Music" Analogy

Raw EEG = Noisy Radio: Lots of static mixed with the signal
Filtering = Noise Cancellation: Removes the "hiss" to hear the music clearly
Frequency Analysis = Music Equalizer: Shows which "notes" (frequencies) are
loudest
Multiple Channels = Orchestra: Each electrode is like a different instrument
📊
## About Our Data

We're using real brain recordings from people doing a memory task:
64 "microphones" placed on the scalp (called electrodes)
500 measurements per second (like a very fast camera for brain activity)
Safe & non-invasive (just like wearing a swimming cap with sensors)
💡 Don't worry about the technical details - we'll explain everything step by step!
🧠
## Assignment 1: EEG Data Loading and Basic Filtering

Think of this like getting your art supplies before painting! We need to gather our "digital
tools" for brain data analysis.
🛠
## Step 1: Getting Your Toolbox Ready


9/27/25, 11:40 AM assignment1_eeg_filtering.ipynb - Colab
https://colab.research.google.com/github/kylemath/NeuroimagingClass/blob/main/assignment1_eeg_filtering.ipynb#scrollTo=A5PFPYJYQyfv&printMo… 1/11

What We're Installing:

MNE-Python 🧠 : The brain data expert (like Photoshop, but for brain signals)
Matplotlib 🎨 : Our artist for making beautiful plots
NumPy 🔢 : The math wizard that does calculations super fast
SciPy ⚗ : Extra science tools for advanced analysis
📱 Mobile/Tablet Users: This might take 1-2 minutes the first time - perfect time for a
coffee break! ☕
# 🎯 STEP 1: Import our digital toolbox!
# Think of this like getting your cooking utensils before making a meal
print ( " 🔧 Setting up your brain data analysis toolkit..." )
print ( " 📱 Mobile users: This is the perfect time to grab some water! 💧 " )
# Check if we're in Google Colab and install packa ges if needed
try :
import google.colab
print ( " 🤖 Detected Google Colab - installing packages..." )
%pip install -q mne
print ( " ✅ Packages installed!" )
except ImportError:
print ( " 🌐 Running in Binder or local environment - packages should alre
# Import our "tools" (like getting your paintbrush es ready)
import mne # 🧠 The brain data expert
import numpy as np # 🔢 The math wizard
import matplotlib.pyplot as plt # 🎨 Our plotting artist
from scipy import signal # ⚗ Extra science tools
import requests # 📡 For downloading data
import os # 📁 File management helper
from pathlib import Path # 🗂 Better file handling
# Make our plots look beautiful and mobile-friendl y
plt.rcParams[ 'figure.figsize' ] = ( 10 , 6 ) # Good size for phones/tablets
plt.rcParams[ 'font.size' ] = 11 # Readable text size
plt.rcParams[ 'lines.linewidth' ] = 1.5 # Nice thick lines
mne.set_log_level( 'WARNING' ) # Less technical chatter
print ( " ✅ SUCCESS! All tools loaded and ready to go!" )
print ( f " 🧠 MNE-Python version: {mne. __version__ } " )
print ( f " 🔢 NumPy version: {np. __version__ } " )
print ( " 🎉 You're all set to analyze some brain waves! 🎉 " )
# 🤖 Google Colab AI Tip:
9/27/25, 11:40 AM assignment1_eeg_filtering.ipynb - Colab
https://colab.research.google.com/github/kylemath/NeuroimagingClass/blob/main/assignment1_eeg_filtering.ipynb#scrollTo=A5PFPYJYQyfv&printMo… 2/11

Click the " ✨ " button next to any code cell to get AI help! # You can ask questions like "What does this code do?" or "How do I modify 🔧 Setting up your brain data analysis toolkit... 📱 Mobile users: This is the perfect time to grab some water! 💧 🤖 Detected Google Colab - installing packages... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 7.4/7 ✅ Packages installed! ✅ SUCCESS! All tools loaded and ready to go! 🧠 MNE-Python version: 1.10.1 🔢 NumPy version: 2.0.2 🎉 You're all set to analyze some brain waves! 🎉 Now for the exciting part - getting some real brain data! This is like downloading a song, but instead of music, we're getting brain wave recordings. 🎵 ## What We're Loading:

Real EEG data from the MNE sample dataset (built-in, no downloads needed!)
Multiple "channels" (like having multiple microphones recording different parts of the
brain)
Continuous recording (like a very long song of brain activity)
🔍
## What You'll See:

Lots of squiggly lines (that's normal - brains are busy!)
Different patterns (some fast wiggles, some slow waves)
Multiple colors (each representing a different electrode location)
📱 Pro Tip: On mobile, you can pinch-to-zoom on any plot to see details better!
📥
## Step 2: Loading Your Brain Wave "Music File"

# 🎯 STEP 2: Load some real brain wave data!
# This is like opening a music file, but instead o f songs, we get brain sig
print ( " 📥 Loading brain wave data (this might take a moment)..." )
print ( " ☕ Perfect time to stretch if you're on mobile!" )
# Get the sample dataset (automatically downloads the first time)
# Think of this like having Spotify download a son g for offline listening
sample_data_folder = mne.datasets.sample.data_path ()
# Load the actual brain recording file
# This is like clicking "play" on your brain wave "song"
raw_file = sample_data_folder / 'MEG' / 'sample' / 'sample_audvis_raw.fif'
raw = mne.io.read_raw_fif(raw_file, preload= True )
9/27/25, 11:40 AM assignment1_eeg_filtering.ipynb - Colab
https://colab.research.google.com/github/kylemath/NeuroimagingClass/blob/main/assignment1_eeg_filtering.ipynb#scrollTo=A5PFPYJYQyfv&printMo… 3/11

Let's pick just the EEG channels (the brain wave "instruments") # This removes other types of sensors we don't nee d for this assignment raw.pick_types(meg= False , eeg= True , stim= False , exclude= 'bads' ) print ( " ✅ SUCCESS! Brain data loaded!" ) print ( f " 📊 Data info:" ) print ( f " 🎵 Length: {raw.times[ -1 ] :.1f } seconds of brain activity" ) print ( f " 🎤 Channels: { len (raw.ch_names)} EEG electrodes" ) print ( f " ⚡ Sampling rate: {raw.info[ 'sfreq' ] :.0f } measurements per secon print ( " 🧠 Ready to explore some brain waves! 🧠 " ) 📥 Loading brain wave data (this might take a moment)... ☕ Perfect time to stretch if you're on mobile! ✅ SUCCESS! Brain data loaded! 📊 Data info: 🎵 Length: 277.7 seconds of brain activity 🎤 Channels: 59 EEG electrodes ⚡ Sampling rate: 601 measurements per second 🧠 Ready to explore some brain waves! 🧠 # 🎯 STEP 3: Visualize the raw brain wave data # This is like looking at the squiggly lines of a song's waveform print ( " 📊 Plotting the raw brain wave data..." ) print ( " 🧐 Examine the squiggly lines - each color is a different electrode! # Plot the raw data for the first 10 seconds from the first 10 channels # We'll limit the number of channels and time to m ake it easier to see raw.plot(duration= 10 , n_channels= 10 , scalings= 'auto' , butterfly= False ) plt.suptitle( 'Raw EEG Data (First 10 Channels, First 10 Seconds )' , y= 1.05 ) plt.tight_layout() plt.show() print ( " ✅ Plot generated! Take a look at the brain waves above." ) print ( " 💡 Notice the different patterns and amplitudes across channels." ) 9/27/25, 11:40 AM assignment1_eeg_filtering.ipynb - Colab https://colab.research.google.com/github/kylemath/NeuroimagingClass/blob/main/assignment1_eeg_filtering.ipynb#scrollTo=A5PFPYJYQyfv&printMo… 4/11

📊 Plotting the raw brain wave data...
🧐 Examine the squiggly lines - each color is a different electrode!

Ella Thorsen-Bell (1803752)
• Note: Code is pasted in document after write-up section (pg 3)
Application Type: Neurofeedback Training Tool Concept Explanation This visual animation illustrates planetary orbits similar to a solar-system. Each “planet” is represented by one EEG band (Delta, Theta, Alpha, Beta, or Gamma). The sun in the middle provides dynamic feedback through threshold-controlled pulsing and colour changing cues. When the sun begins to glow and pulse with a green hue, it represents focus within the user. In contrast, if the user gets distracted and loses focus, the sun turns a pulsing red colour. The ring around the Beta wave planet also emphasizes the focus state, providing an easy cue to visualize and better understand patterns. Each addition to this application relates to the overall space theme in order to maintain cohesiveness and create a visual polish. This application is meant to serve as an easy tool to track mental engagement, as well as help improve focus, self-awareness, and control. This tool can also be altered to track and train other brain states through small adjustments in code. Technical Implementation This application uses real-time readings of an individual’s brainwave frequencies and attention level to influence a variety of factors within the animation. More specifically it uses a combination of codes eegData.alpha, .beta, .theta, .delta, .gamma, & .attention. I used the “Band Orbitz” template as a base for this assignment, where each EEG frequency band is represented as a planet in orbit with a specific radius and colour. The values of the data impact the size and speed of each planet. The most significant change to this template related to the overall goal of this application: the addition of neurofeedback signals. In order to achieve this, I added the yellow shape and a glow effect to the center using “fill(), ellipse() & glowSize” functions before

manipulating an overlaying translucent colour and pulsing pattern on top with the code. The
threshold for this tool was set to 0.7. An attention level over 0.7 is indicated by a slow pulse that
expands and contracts from the center, whereas an attention level under 0.7 is indicated by a
faster red pulse. This was added using fill(), ellipse() & stroke(). The Beta frequency band is also
highlighted through the use of a ring using noFill(), stroke (), ellipse() functions to both add to
the overall theme & bring attention to the specific brain state. I also altered the radius of each
orbit to fit better on the screen and added a starry background.
Challenges & What I Learned
One challenge I faced related to the direct understanding of how to work with certain
aspects of code. Throughout this assignment I went through a lot of trial and error, learning how
to manipulate and add to existing code, as well as work through error messages. I did a number
of searches to add a variety of features since I’m not the most familiar with coding. Another
significant challenge was figuring out how to substantially change the code. I spent time testing
out different applications and altering smaller elements but had to learn how to build upon them
in order to create something bigger. For example, I tried to manipulate the “alpha wave” template
into the brain art by manipulating the colour to warm/cool colours based on alpha and theta, but
it didn’t feel like enough. I also learned more about how neurofeedback training tools work,
since I had to figure out what to incorporate into my own neurofeedback application.

Code:
// Solar System Orbits - frequency bands
let angle = 0 ;
let stars = [];
function setup() {
colorMode( HSB , 360 , 100 , 100 );
// Create random stars
for ( let i = 0 ; i < 300 ; i++) {
stars.push({
x: random(-width, width),
y: random(-height, height),
size: random( 1 , 3 ),
brightness: random( 60 , 100 )
});
}
}
function draw() {
background( 240 , 50 , 5 , 0.2 );
// Draw star field
noStroke();
for ( let s of stars) {
fill( 60 , 10 , s.brightness, 0.6 );
ellipse(s.x, s.y, s.size);
}
translate(width/ 2 , height/ 2 );
//Focus threshold
let isFocused = eegData.attention > 0.7 ;
let bands = [
{ name: 'Delta' , value: eegData.delta, hue: 10 , sat: 40 , bright: 60 , radius: 50 },
{ name: 'Theta' , value: eegData.theta, hue: 30 , sat: 60 , bright: 70 , radius: 90 },
{ name: 'Alpha' , value: eegData.alpha, hue: 140 , sat: 50 , bright: 60 , radius: 130 },
{ name: 'Beta' , value: eegData.beta, hue: 200 , sat: 70 , bright: 20 , radius: 170 },
{ name: 'Gamma' , value: eegData.gamma, hue: 300 , sat: 50 , bright: 80 , radius: 210 }
];
for ( let i = 0 ; i < bands.length; i++) {
let band = bands[i];
let a = angle + (i * TWO_PI / bands.length);
let r = band.radius;
let size = 10 + band.value * 40 ;

let x = cos(a) * r;
let y = sin(a) * r;
//Add rings to Beta
fill(band.hue, 70 , 90 , 0.8 );
noStroke();
ellipse(x, y, size);
if (band.name === 'Beta' ) {
noFill();
stroke(band.hue, 50 , 100 , 0.5 );
ellipse(x, y, size * 1.5 , size * 0.7 ); // Ring
}
// Draw orbit path
noFill();
stroke(band.hue, 20 , 60 , 0.1 );
ellipse( 0 , 0 , r * 2 );
}
// Draw sun in center
fill( 50 , 80 , 100 , 1 );
ellipse( 0 , 0 , 45 );
for ( let i = 5 ; i > 0 ; i--) {
let glowSize = 15 + i * 10 ;
fill( 50 , 80 , 100 , 0.05 * i);
noStroke();
ellipse( 0 , 0 , glowSize);
}
//Neurofeedback signals
if (isFocused) {
fill( 120 , 60 , 100 , 0.2 );
ellipse( 0 , 0 , 75 + sin(frameCount * 0.1 ) * 10 );
} else {
if (!isFocused) {
fill( 0 , 80 , 100 , 0.15 );
stroke( 0 , 80 , 100 , 0.1 + 0.1 * sin(frameCount * 0.3 ));
ellipse( 0 , 0 , 80 + sin(frameCount * 0.3 ) * 6 );
}
}
angle += 0.02 * eegData.attention;
}

Ella Thorsen-Bell (1803752)
Final Project Selection (Option 1: Mini Data Analysis)
For the final project I plan to do a mini data analysis based on the following research
question: How do sensory evoked potentials (SEPs) differ between auditory and visual stimuli?
In order to respond to the question, I will compare the ERP’s for both categories (visual vs
auditory). I also plan to compare the amplitude of certain electrodes such as N1, P2, N75, &
P100 if available in the dataset(s). I will then use this to compare the peak times across channels.
If available, I may also use a topographical distribution analysis. This will allow me to look at
both the difference in timing as well as the brain regions involved. Understanding these
differences could aid in providing more precise diagnostic tools or experimental designs.

Ella Thorsen-Bell (1803752) Part 1: BCI Innovation: Beyond Basic Control (Option B) For this project I implemented a BCI that combines real-time EEG signals with auditory feedback to create a unique soundscape, specifically using alpha, beta, & theta frequency band oscillators. It works through a continuous feedback loop, where EEG signals are read and mapped into specific sound and visual properties. Alpha activity is translated into lower pitched sounds using sine waves, which is also represented through the animation on screen. Beta activity controls the higher pitches using square-waves and influences the speed of the animation. Theta waves are expressed through a subtle bass layer to represent deep relaxation. Reverb and delay effects were also added to enrich the soundscape. All of them are layered into one simultaneous presentation. This application could be used for a variety of purposes such as neurofeedback, brain art, or potentially even research. This application is different from basic parameter control due to the layered auditory feedback. Instead of mapping a single parameter or only expressing a single frequency band, it attempts to combine multiple frequency bands into one simultaneous auditory environment. Throughout the process I learned how to adjust different frequencies to incoming EEG signals, including how to adjust the amplitude and pitch of each frequency band. I also learned more generally how to implement sound into BCI systems, allowing me to enrich my chosen application. With more manipulations in code, this application has the potential to grow further through the addition of better visual feedback to accompany the auditory experience. Video: ThorsenBell_Midterm1_Part1Vid.webm
• Note: Video recorded using simulated data, real data may appear differently
• Sound needs to be on to work but careful of the volume (it can be loud)

Part 2: Event-Related Potentials: Build Your Own ERP Experiment For this experiment, I focused on measuring the P300 event related potential (ERP), which is linked to cognitive processes such as attention, working memory, and decision making. The stimulus manipulation of this experiment follows an oddball paradigm using two distinct categories of stimuli: frequent and rare. Each stimulus appears at a different probability, with the frequent stimulus presenting 80% of the time and the rare stimulus appearing only 20% of the time. Both stimuli are shown in 30ms intervals with a 60ms gap. There is also a 1000ms interval between each trial. For this experiment an averaging system is used to isolate the brain’s response to a specific event, or in this case the presentation of the rare stimulus. It works by continuously recording EEG data & extracting time-locked epochs. The epochs are then averaged separately for each stimulus category to allow for independent averaging and to generate the corresponding ERP wavelengths. Based on existing literature on P300 ERP oddball experiments, I would expect to see a larger P300 response to the rare stimulus in comparison to the frequent one, due to the rare one being less predictable to the brain. A positive peak around 300ms post-stimulus onset would also be expected, since that amplitude is particularly sensitive to novelty. Video: ThorsenBell_Midterm_Part2Vid.webm
• Note: Video recorded using simulated data, real data may appear differently