Explore how spectral power and connectivity features interact with machine-learning models across seven psychiatric classes. Adjust the EEG profile, select a model and run an educational simulation based on the study results.
945Participants
1,140EEG features
7Classification classes
91.57%Theta standalone accuracy
Live EEG profile
The waveform reacts as you change the spectral sliders.
Profile view: Move the sliders to explore how the schematic waveform changes. This is an educational visualization, not a patient diagnosis.
Study performance
Support Vector Machine · Obsessive–Compulsive Disorder
Precision94.56%
Recall93.45%
F1-score94.00%
Ready. Choose a psychiatric class and machine-learning model, then run the analysis.
READY TO TEST YOURSELF?
Turn the article into a research challenge.
Compare classifiers, explore EEG features and rebuild the study pipeline.
Educational research explorer · not a clinical diagnostic tool
7 ROUNDS · ARTICLE DATA
Can You Beat the Model?
Choose the classifier with the highest reported F1-score for each diagnostic class.
ROUND 1 / 7STREAK × 0
Which classifier achieved the highest F1-score?
Choose a model.
ROUND SCORE
0 XP
Reported F1-scores
Values appear after your answer.
INTERACTIVE VISUALISATION
EEG Profile Simulator
Manipulate the retained frequency bands and switch between schematic power and connectivity views.
Spectral controls
SIMULATED INTERPRETATION
—
Simulation match—
Adjust the bands and press ANALYSE PROFILE. The resulting class is a game-like simulated output, not a clinical diagnosis.
Spectral Power View
INTERPRETABILITY · STEP-BY-STEP
Interactive EEG Feature Explorer
See how the study transformed EEG information into features used by the machine-learning models.
1
Choose a frequency bandTheta, Alpha, Beta or High-Beta
→
2
Choose what to explorePower, connectivity or top features
→
3
Interact with the EEG mapClick one or two electrodes as instructed
STEP 1 · FREQUENCY BAND
STEP 2 · FEATURE TYPE
STEP 3 · EEG MAP
Click one electrode to create an absolute-power feature.
Available electrode Your selection Study-highlighted region
WHAT ARE YOU BUILDING?
Absolute Power
Choose a band and click one electrode. The app will show how a channel-level spectral-power feature is represented.
YOUR FEATURE
Select an electrodeStart by clicking one electrode on the EEG map.
WHAT DOES THIS MEAN?
Absolute power describes the strength of EEG activity within the selected frequency band at one electrode.
19absolute-power features per band
171pairwise coherence features per band
190EEG descriptors per band
Try it: keep Alpha selected and click Pz to see a feature related to the region highlighted in the study.
ARTICLE CONNECTION
Why are these electrodes highlighted?
The study interpreted feature importance using absolute weights from the linear Support Vector Machine. Alpha-band parietal and occipital features were prominent, and AB_C_alpha_o_Pz was reported as the top-ranked feature.
P3PzP4O1O2
Mini challenge · Find Pz
After exploring the map, test yourself: can you identify the parietal midline electrode associated with the reported top-ranked alpha feature?
REPORTED RESULTS
Results Dashboard
Compare testing accuracy, precision, recall and F1-score without mixing the metrics.
BEST REPORTED F1 BY CLASS
METHODS GAME
Build the Study Pipeline
Arrange the implemented analysis stages, then identify the preprocessing caveat acknowledged by the authors.
Use ↑ and ↓ to reorder the stages.
CRITICAL THINKING
Spot the methodological issue
When were feature ranking and z-score normalisation performed in the implemented analysis?
Choose the workflow reported in the article.
5 QUESTIONS
Final BRAIN Research Challenge
Finish with questions on the dataset, EEG bands, interpretation and validation design.
QUESTION 1 / 50 / 500 XP
Choose one answer.
★
CHALLENGE COMPLETE
TOTAL 0 XP
MINI CHALLENGE
Can you find Pz?
Pz is the parietal midline electrode. Click the correct position on the simplified EEG map below.
