neurology EEG Detection

Live EEG Signal Visualization

400Hz Sampling 5 Channels
C3 Cz C4 P3 P4
cloud_upload

Upload Raw EEG Data

Compatible: .edf, .csv, .mat, .npy

check_circle Live Connection Active

Neural Classification Output

87.6% CONFIDENCE
Left Hand Imagery
Left Hand Projection82%
Right Hand Projection18%
database

PhysioNet EEG Dataset Specs

Subjects 109 Patients
Channels 64 Array
States 4 Imagery
memory

Architecture: Deep-ANN

Active Model
Optimizer Adam (lr=0.001)
Activation Leaky ReLU
Loss Function Categorical Crossentropy
Training Epochs 250 (Early Stop)
Dropout Rate 0.45 Rate
Output Layer Softmax

Neural Processing Pipeline

waves Raw EEG Input
double_arrow
filter_alt Filtering (BPF)
double_arrow
cleaning_services Artifact Removal
double_arrow
biotech Feature Extraction
double_arrow
psychology ANN Classification

Extracted Spectral Features

Signal Mean RMS Amplitude Alpha Power (8-13Hz) Beta Power (13-30Hz) Spectral Entropy Hurst Exponent PSD Estimation Hjorth Parameters