Python结合TensorFlow构建U型网络（Unet）用于医学图像分割是一个常见的深度学习应用。U型网络因其递归结构和上采样能力，在处理像病理切片、MRI等高分辨率图像时非常有效。以下是使用Keras库（它是TensorFlow的一个高级API）创建一个基础版U型网络的简化代码示例： ```python import tensorflow as tf from tensorflow.keras.models import Model from tensorflow.keras.layers import Input, Conv2D, MaxPooling2D, UpSampling2D, Concatenate def unet(input_size=(256, 256, 1), num_classes=1): inputs = Input(input_size) # 编码阶段 conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(inputs) conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv1) pool1 = MaxPooling2D(pool_size=(2, 2))(conv1) conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(pool1) conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv2) pool2 = MaxPooling2D(pool_size=(2, 2))(conv2) # 中间层 conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(pool2) conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(conv3) # 解码阶段 up4 = Conv2D(64, (2, 2), activation='relu', padding='same')(UpSampling2D(size=(2, 2))(conv3)) merge4 = Concatenate()([conv2, up4]) conv4 = Conv2D(64, (3, 3), activation='relu', padding='same')(merge4) conv4 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv4) up5 = Conv2D(32, (2, 2), activation='relu', padding='same')(UpSampling2D(size=(2, 2))(conv4)) merge5 = Concatenate()([conv1, up5]) conv5 = Conv2D(32, (3, 3), activation='relu', padding='same')(merge5) conv5 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv5) # 输出层 output = Conv2D(num_classes, (1, 1), activation='sigmoid')(conv5) model = Model(inputs=inputs, outputs=output) return model # 创建并编译模型 model = unet() model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy']) # 这里只是一个基本模板，实际使用时你需要准备训练数据、调整超参数，并进行训练 ```