脑机接口深度解析：EEG 信号处理核心算法与 Python 实战

脑机接口深度解析：EEG 信号处理核心算法与 Python 实战

1. 技术分析

1.1 脑机接口概述

脑机接口(BCI)是连接大脑和外部设备的技术：

BCI分类
侵入式: 植入大脑内部
非侵入式: 外部设备
半侵入式: 部分植入

BCI应用:
医疗康复
增强认知
娱乐体验
神经控制

1.2 BCI工作原理

BCI工作流程
信号采集: EEG/ECoG/fMRI
信号处理: 滤波、降噪
特征提取: 提取脑电特征
模式识别: 识别意图
动作执行: 控制外部设备

信号类型:
EEG: 脑电图
ECoG: 皮层脑电图
fMRI: 功能性磁共振

1.3 BCI技术挑战

技术挑战
信号质量: 噪声干扰
空间分辨率: 精度有限
长期稳定性: 植入物寿命
生物相容性: 排异反应

研究方向:
新材料
无线传输
机器学习
神经修复

2. 核心功能实现

2.1 EEG信号处理

import numpy as np

class EEGProcessor:
def __init__(self, sampling_rate=256):
self.sampling_rate = sampling_rate

def load_signal(self, file_path):
return np.load(file_path)

def filter_signal(self, signal, low_freq=1, high_freq=50):
nyquist = 0.5 * self.sampling_rate

low = low_freq / nyquist
high = high_freq / nyquist

b, a = self._butter_bandpass(low, high, order=4)
filtered = self._apply_filter(signal, b, a)

return filtered

def _butter_bandpass(self, low, high, order=4):
from scipy.signal import butter
return butter(order, [low, high], btype='band')

def _apply_filter(self, signal, b, a):
from scipy.signal import lfilter
return lfilter(b, a, signal)

def extract_features(self, signal, window_size=128):
features = []

for i in range(0, len(signal) – window_size, window_size):
window = signal[i:i+window_size]

mean = np.mean(window)
std = np.std(window)
peak_to_peak = np.max(window) – np.min(window)

features.append([mean, std, peak_to_peak])

return np.array(features)

def detect_motor_intent(self, features):
model = self._load_model()
prediction = model.predict(features)

intent_map = {0: 'rest', 1: 'left_hand', 2: 'right_hand', 3: 'foot'}
return intent_map.get(prediction[-1], 'unknown')

def _load_model(self):
from sklearn.linear_model import LogisticRegression
return LogisticRegression()

2.2 神经解码器

class NeuralDecoder:
def __init__(self):
self.decoders = {}

def train_decoder(self, decoder_id, training_data):
X = training_data['features']
y = training_data['labels']

from sklearn.svm import SVC
model = SVC(kernel='rbf', C=1.0)
model.fit(X, y)

self.decoders[decoder_id] = model

def decode(self, decoder_id, features):
if decoder_id not in self.decoders:
raise ValueError("Decoder not found")

model = self.decoders[decoder_id]
prediction = model.predict(features)

return prediction

def evaluate_decoder(self, decoder_id, test_data):
model = self.decoders.get(decoder_id)

if not model:
return None

X = test_data['features']
y = test_data['labels']

accuracy = model.score(X, y)
return accuracy

2.3 BCI控制器

class BCIController:
def __init__(self):
self.devices = {}

def register_device(self, device_id, device_type):
self.devices[device_id] = {
'type': device_type,
'status': 'idle',
'position': (0, 0, 0)
}

def control_device(self, device_id, command):
if device_id not in self.devices:
return False

device = self.devices[device_id]

if device['type'] == 'cursor':
self._move_cursor(device_id, command)
elif device['type'] == 'robot_arm':
self._move_robot_arm(device_id, command)

return True

def _move_cursor(self, device_id, command):
current_pos = self.devices[device_id]['position']

if command == 'up':
new_pos = (current_pos[0], current_pos[1] – 10, current_pos[2])
elif command == 'down':
new_pos = (current_pos[0], current_pos[1] + 10, current_pos[2])
elif command == 'left':
new_pos = (current_pos[0] – 10, current_pos[1], current_pos[2])
elif command == 'right':
new_pos = (current_pos[0] + 10, current_pos[1], current_pos[2])
else:
new_pos = current_pos

self.devices[device_id]['position'] = new_pos

def _move_robot_arm(self, device_id, command):
print(f"Moving robot arm: {command}")

3. 性能对比

3.1 BCI类型对比

3.2 信号质量对比

3.3 BCI应用对比

4. 最佳实践

4.1 EEG信号处理示例

def eeg_processing_example():
processor = EEGProcessor()

signal = np.random.randn(2560)

filtered = processor.filter_signal(signal)
features = processor.extract_features(filtered)

intent = processor.detect_motor_intent(features)
print(f"Detected intent: {intent}")

4.2 BCI控制示例

def bci_control_example():
controller = BCIController()

controller.register_device('cursor', 'cursor')
controller.register_device('arm', 'robot_arm')

commands = ['up', 'right', 'down', 'left']

for cmd in commands:
success = controller.control_device('cursor', cmd)
print(f"Control {cmd}: {success}")

pos = controller.devices['cursor']['position']
print(f"Cursor position: {pos}")

5. 总结

脑机接口技术正在突破人机边界：

对比数据如下：

• EEG最便携
• ECoG平衡性能
• 侵入式最精确
• 医疗康复最成熟

脑机接口将在医疗、娱乐、增强认知等领域带来革命性变化。