Medical IoT and AI

Wearable Health Data Processing

from scipy.signal import butter, filtfilt
import numpy as np

class WearableDataProcessor:
    def __init__(self, sampling_rate=250):
        self.fs = sampling_rate

    def bandpass_filter(self, signal, lowcut=0.5, highcut=40.0, order=4):
        nyq = 0.5 * self.fs
        b, a = butter(order, [lowcut/nyq, highcut/nyq], btype='band')
        return filtfilt(b, a, signal)

    def normalize(self, signal):
        return (signal - np.mean(signal)) / (np.std(signal) + 1e-8)

Anomaly Detection in IoT Streams

Statistical Process Control

LSTM-Based Detection

class IoTAnomalyDetector(nn.Module):
    def __init__(self, n_features, hidden_size=64):
        super().__init__()
        self.lstm = nn.LSTM(n_features, hidden_size, batch_first=True)
        self.decoder = nn.Sequential(
            nn.Linear(hidden_size, n_features * 2), nn.ReLU(),
            nn.Linear(n_features * 2, n_features))

    def forward(self, x):
        _, (h, _) = self.lstm(x)
        return self.decoder(h.squeeze(0))

    def anomaly_score(self, x):
        return torch.mean((x - self.forward(x)) ** 2, dim=-1)

Edge Computing for Health

Model Compression

Quantization