Prosthetics Control

EMG Signal Classification

Feature Extraction

import numpy as np

class EMGFeatureExtractor:
    def extract(self, signal):
        return {
            'MAV': np.mean(np.abs(signal)),
            'WL': np.sum(np.abs(np.diff(signal))),
            'ZC': np.sum(np.diff(np.sign(signal)) != 0),
            'RMS': np.sqrt(np.mean(signal ** 2)),
            'MF': self._mean_frequency(signal),
            'MNF': self._mean_freq(signal)
        }

    def _mean_frequency(self, signal):
        fft = np.abs(np.fft.rfft(signal))
        freqs = np.fft.rfftfreq(len(signal))
        return np.sum(freqs * fft) / (np.sum(fft) + 1e-8)

Intent Recognition

class IntentClassifier(nn.Module):
    def __init__(self, n_channels=8, n_classes=7):
        super().__init__()
        self.conv = nn.Sequential(
            nn.Conv1d(n_channels, 32, 5, padding=2), nn.ReLU(), nn.MaxPool1d(2),
            nn.Conv1d(32, 64, 5, padding=2), nn.ReLU(), nn.AdaptiveAvgPool1d(1))
        self.fc = nn.Linear(64, n_classes)

    def forward(self, x):
        return self.fc(self.conv(x).squeeze(-1))

Adaptive Prosthetic Control

Co-contraction Detection