活动地址：CSDN21天学习挑战赛
- 🍨 本文为🔗365天深度学习训练营 中的学习记录博客
- 🍦 参考文章地址： 🔗深度学习100例 | 第26天-卷积神经网络（CNN）：乳腺癌识别
- 🍖 作者：K同学啊
今天继续学习CNN，案例是深度学习在医学领域的应用，乳腺癌是女性最常见的癌症形式，浸润性导管癌 (IDC) 是最常见的乳腺癌形式。准确识别和分类乳腺癌亚型是一项重要的临床任务，利用深度学习方法识别可以有效节省时间并减少错误。(完整源码附后)
1、数据导入、配置
数据集是由多张以 40 倍扫描的乳腺癌 (BCa) 标本的完整载玻片图像组成。
分为两类，即：正常细胞、乳腺癌细胞
图片总数为： 13403
1.1 导入数据
import matplotlib.pyplot as plt
import os,PIL,pathlib
import numpy as np
import pandas as pd
import warnings
from tensorflow import keras
import pathlib
warnings.filterwarnings("ignore")
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False
data_dir = "./26-data"
data_dir = pathlib.Path(data_dir)
image_count = len(list(data_dir.glob('*/*')))
print("图片总数为：",image_count)
batch_size = 16
img_height = 50
img_width = 50
train_ds = tf.keras.preprocessing.image_dataset_from_directory(
data_dir,
validation_split=0.2,
subset="training",
seed=12,
image_size=(img_height, img_width),
batch_size=batch_size)
val_ds = tf.keras.preprocessing.image_dataset_from_directory(
data_dir,
validation_split=0.2,
subset="validation",
seed=12,
image_size=(img_height, img_width),
batch_size=batch_size)
class_names = train_ds.class_names
print(class_names)
for image_batch, labels_batch in train_ds:
print(image_batch.shape)
print(labels_batch.shape)
break
1.2 配置数据集
AUTOTUNE = tf.data.experimental.AUTOTUNE
def train_preprocessing(image,label):
return (image/255.0,label)
train_ds = (
train_ds.cache()
.shuffle(1000)
.map(train_preprocessing) # 这里可以设置预处理函数
# .batch(batch_size) # 在image_dataset_from_directory处已经设置了batch_size
.prefetch(buffer_size=AUTOTUNE)
)
val_ds = (
val_ds.cache()
.shuffle(1000)
.map(train_preprocessing) # 这里可以设置预处理函数
# .batch(batch_size) # 在image_dataset_from_directory处已经设置了batch_size
.prefetch(buffer_size=AUTOTUNE)
)
1.3 可视化预览数据
plt.figure(figsize=(10, 8)) # 图形的宽为10高为5
plt.suptitle("数据展示")
class_names = ["乳腺癌细胞","正常细胞"]
for images, labels in train_ds.take(1):
for i in range(15):
plt.subplot(4, 5, i + 1)
plt.xticks([])
plt.yticks([])
plt.grid(False)
# 显示图片
plt.imshow(images[i])
# 显示标签
plt.xlabel(class_names[labels[i]-1])
plt.show()
2、模型的构建、编译、训练
2.1 构建模型
model = tf.keras.Sequential([
tf.keras.layers.Conv2D(filters=16,kernel_size=(3,3),padding="same",activation="relu",input_shape=[img_width, img_height, 3]),
tf.keras.layers.Conv2D(filters=16,kernel_size=(3,3),padding="same",activation="relu"),
tf.keras.layers.MaxPooling2D((2,2)),
tf.keras.layers.Dropout(0.5),
tf.keras.layers.Conv2D(filters=16,kernel_size=(3,3),padding="same",activation="relu"),
tf.keras.layers.MaxPooling2D((2,2)),
tf.keras.layers.Conv2D(filters=16,kernel_size=(3,3),padding="same",activation="relu"),
tf.keras.layers.MaxPooling2D((2,2)),
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(2, activation="softmax")
])
model.summary()
2.2 编译模型
model.compile(optimizer="adam",
loss='sparse_categorical_crossentropy',
metrics=['accuracy'])
2.3 训练模型
from tensorflow.keras.callbacks import ModelCheckpoint, Callback, EarlyStopping, ReduceLROnPlateau, LearningRateScheduler
NO_EPOCHS = 100
PATIENCE = 5
VERBOSE = 1
# 设置动态学习率
annealer = LearningRateScheduler(lambda x: 1e-3 * 0.99 ** (x+NO_EPOCHS))
# 设置早停
earlystopper = EarlyStopping(monitor='loss', patience=PATIENCE, verbose=VERBOSE)
#
checkpointer = ModelCheckpoint('best_model.h5',
monitor='val_accuracy',
verbose=VERBOSE,
save_best_only=True,
save_weights_only=True)
train_model = model.fit(train_ds,
epochs=NO_EPOCHS,
verbose=1,
validation_data=val_ds,
callbacks=[earlystopper, checkpointer, annealer])
训练结果：
Epoch 90/100
670/671 [============================>.] - ETA: 0s - loss: 0.2175 - accuracy: 0.9089
Epoch 00090: val_accuracy did not improve from 0.89851
671/671 [==============================] - 19s 29ms/step - loss: 0.2174 - accuracy: 0.9089 - val_loss: 0.3031 - val_accuracy: 0.8649
Epoch 91/100
670/671 [============================>.] - ETA: 0s - loss: 0.2185 - accuracy: 0.9083
Epoch 00091: val_accuracy did not improve from 0.89851
671/671 [==============================] - 19s 29ms/step - loss: 0.2183 - accuracy: 0.9083 - val_loss: 0.3082 - val_accuracy: 0.8701
Epoch 92/100
671/671 [==============================] - ETA: 0s - loss: 0.2160 - accuracy: 0.9074
Epoch 00092: val_accuracy did not improve from 0.89851
671/671 [==============================] - 18s 27ms/step - loss: 0.2160 - accuracy: 0.9074 - val_loss: 0.2741 - val_accuracy: 0.8869
Epoch 93/100
671/671 [==============================] - ETA: 0s - loss: 0.2165 - accuracy: 0.9103
Epoch 00093: val_accuracy did not improve from 0.89851
671/671 [==============================] - 20s 30ms/step - loss: 0.2165 - accuracy: 0.9103 - val_loss: 0.2739 - val_accuracy: 0.8877
Epoch 94/100
671/671 [==============================] - ETA: 0s - loss: 0.2152 - accuracy: 0.9080
Epoch 00094: val_accuracy did not improve from 0.89851
671/671 [==============================] - 18s 27ms/step - loss: 0.2152 - accuracy: 0.9080 - val_loss: 0.2740 - val_accuracy: 0.8866
Epoch 95/100
671/671 [==============================] - ETA: 0s - loss: 0.2138 - accuracy: 0.9092
Epoch 00095: val_accuracy did not improve from 0.89851
671/671 [==============================] - 18s 27ms/step - loss: 0.2138 - accuracy: 0.9092 - val_loss: 0.3400 - val_accuracy: 0.8541
Epoch 96/100
671/671 [==============================] - ETA: 0s - loss: 0.2149 - accuracy: 0.9092
Epoch 00096: val_accuracy did not improve from 0.89851
671/671 [==============================] - 18s 26ms/step - loss: 0.2149 - accuracy: 0.9092 - val_loss: 0.2897 - val_accuracy: 0.8806
Epoch 97/100
670/671 [============================>.] - ETA: 0s - loss: 0.2123 - accuracy: 0.9110
Epoch 00097: val_accuracy did not improve from 0.89851
671/671 [==============================] - 18s 27ms/step - loss: 0.2122 - accuracy: 0.9110 - val_loss: 0.3222 - val_accuracy: 0.8593
Epoch 98/100
671/671 [==============================] - ETA: 0s - loss: 0.2120 - accuracy: 0.9107
Epoch 00098: val_accuracy did not improve from 0.89851
671/671 [==============================] - 19s 28ms/step - loss: 0.2120 - accuracy: 0.9107 - val_loss: 0.3023 - val_accuracy: 0.8757
Epoch 99/100
670/671 [============================>.] - ETA: 0s - loss: 0.2113 - accuracy: 0.9114
Epoch 00099: val_accuracy did not improve from 0.89851
671/671 [==============================] - 19s 29ms/step - loss: 0.2112 - accuracy: 0.9114 - val_loss: 0.2606 - val_accuracy: 0.8918
Epoch 100/100
670/671 [============================>.] - ETA: 0s - loss: 0.2130 - accuracy: 0.9112
Epoch 00100: val_accuracy did not improve from 0.89851
671/671 [==============================] - 19s 28ms/step - loss: 0.2131 - accuracy: 0.9112 - val_loss: 0.2939 - val_accuracy: 0.8799
3、评估模型
3.1 Accuracy与Loss图
acc = train_model.history['accuracy']
val_acc = train_model.history['val_accuracy']
loss = train_model.history['loss']
val_loss = train_model.history['val_loss']
epochs_range = range(len(acc))
plt.figure(figsize=(12, 4))
plt.subplot(1, 2, 1)
plt.plot(epochs_range, acc, label='Training Accuracy')
plt.plot(epochs_range, val_acc, label='Validation Accuracy')
plt.legend(loc='lower right')
plt.title('Training and Validation Accuracy')
plt.subplot(1, 2, 2)
plt.plot(epochs_range, loss, label='Training Loss')
plt.plot(epochs_range, val_loss, label='Validation Loss')
plt.legend(loc='upper right')
plt.title('Training and Validation Loss')
plt.show()
3.2 混淆矩阵
from sklearn.metrics import confusion_matrix
import seaborn as sns
import pandas as pd
# 定义一个绘制混淆矩阵图的函数
def plot_cm(labels, predictions):
# 生成混淆矩阵
conf_numpy = confusion_matrix(labels, predictions)
# 将矩阵转化为 DataFrame
conf_df = pd.DataFrame(conf_numpy, index=class_names ,columns=class_names)
plt.figure(figsize=(8,7))
sns.heatmap(conf_df, annot=True, fmt="d", cmap="BuPu")
plt.title('混淆矩阵',fontsize=15)
plt.ylabel('真实值',fontsize=14)
plt.xlabel('预测值',fontsize=14)
val_pre = []
val_label = []
for images, labels in val_ds:#这里可以取部分验证数据（.take(1)）生成混淆矩阵
for image, label in zip(images, labels):
# 需要给图片增加一个维度
img_array = tf.expand_dims(image, 0)
# 使用模型预测图片中的人物
prediction = model.predict(img_array)
val_pre.append(class_names[np.argmax(prediction)])
val_label.append(class_names[label])
plot_cm(val_label, val_pre)
3.3 各项指标评估
from sklearn import metrics
def test_accuracy_report(model):
print(metrics.classification_report(val_label, val_pre, target_names=class_names))
score = model.evaluate(val_ds, verbose=0)
print('Loss function: %s, accuracy:' % score[0], score[1])
test_accuracy_report(model)
完整源码
import tensorflow as tf
import matplotlib.pyplot as plt
import os,PIL,pathlib
import numpy as np
import pandas as pd
import warnings
from tensorflow import keras
warnings.filterwarnings("ignore") #忽略警告信息
plt.rcParams['font.sans-serif'] = ['SimHei'] # 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False # 用来正常显示负号
import pathlib
data_dir = "./26-data"
data_dir = pathlib.Path(data_dir)
image_count = len(list(data_dir.glob('*/*')))
print("图片总数为：",image_count)
batch_size = 16
img_height = 50
img_width = 50
"""
关于image_dataset_from_directory()的详细介绍可以参考文章：https://mtyjkh.blog.csdn.net/article/details/117018789
"""
train_ds = tf.keras.preprocessing.image_dataset_from_directory(
data_dir,
validation_split=0.2,
subset="training",
seed=12,
image_size=(img_height, img_width),
batch_size=batch_size)
"""
关于image_dataset_from_directory()的详细介绍可以参考文章：https://mtyjkh.blog.csdn.net/article/details/117018789
"""
val_ds = tf.keras.preprocessing.image_dataset_from_directory(
data_dir,
validation_split=0.2,
subset="validation",
seed=12,
image_size=(img_height, img_width),
batch_size=batch_size)
class_names = train_ds.class_names
print(class_names)
for image_batch, labels_batch in train_ds:
print(image_batch.shape)
print(labels_batch.shape)
break
AUTOTUNE = tf.data.experimental.AUTOTUNE
# 这里如果是 tf2.6 或者报错，使用 AUTOTUNE = tf.data.experimental.AUTOTUNE
def train_preprocessing(image,label):
return (image/255.0,label)
train_ds = (
train_ds.cache()
.shuffle(1000)
.map(train_preprocessing) # 这里可以设置预处理函数
# .batch(batch_size) # 在image_dataset_from_directory处已经设置了batch_size
.prefetch(buffer_size=AUTOTUNE)
)
val_ds = (
val_ds.cache()
.shuffle(1000)
.map(train_preprocessing) # 这里可以设置预处理函数
# .batch(batch_size) # 在image_dataset_from_directory处已经设置了batch_size
.prefetch(buffer_size=AUTOTUNE)
)
plt.figure(figsize=(10, 8)) # 图形的宽为10高为5
plt.suptitle("数据展示")
class_names = ["乳腺癌细胞","正常细胞"]
for images, labels in train_ds.take(1):
for i in range(15):
plt.subplot(4, 5, i + 1)
plt.xticks([])
plt.yticks([])
plt.grid(False)
# 显示图片
plt.imshow(images[i])
# 显示标签
plt.xlabel(class_names[labels[i]-1])
plt.show()
import tensorflow as tf
model = tf.keras.Sequential([
tf.keras.layers.Conv2D(filters=16,kernel_size=(3,3),padding="same",activation="relu",input_shape=[img_width, img_height, 3]),
tf.keras.layers.Conv2D(filters=16,kernel_size=(3,3),padding="same",activation="relu"),
tf.keras.layers.MaxPooling2D((2,2)),
tf.keras.layers.Dropout(0.5),
tf.keras.layers.Conv2D(filters=16,kernel_size=(3,3),padding="same",activation="relu"),
tf.keras.layers.MaxPooling2D((2,2)),
tf.keras.layers.Conv2D(filters=16,kernel_size=(3,3),padding="same",activation="relu"),
tf.keras.layers.MaxPooling2D((2,2)),
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(2, activation="softmax")
])
model.summary()
model.compile(optimizer="adam",
loss='sparse_categorical_crossentropy',
metrics=['accuracy'])
from tensorflow.keras.callbacks import ModelCheckpoint, Callback, EarlyStopping, ReduceLROnPlateau, LearningRateScheduler
NO_EPOCHS = 100
PATIENCE = 5
VERBOSE = 1
# 设置动态学习率
annealer = LearningRateScheduler(lambda x: 1e-3 * 0.99 ** (x+NO_EPOCHS))
# 设置早停
earlystopper = EarlyStopping(monitor='loss', patience=PATIENCE, verbose=VERBOSE)
#
checkpointer = ModelCheckpoint('best_model.h5',
monitor='val_accuracy',
verbose=VERBOSE,
save_best_only=True,
save_weights_only=True)
train_model = model.fit(train_ds,
epochs=NO_EPOCHS,
verbose=1,
validation_data=val_ds,
callbacks=[earlystopper, checkpointer, annealer])
acc = train_model.history['accuracy']
val_acc = train_model.history['val_accuracy']
loss = train_model.history['loss']
val_loss = train_model.history['val_loss']
epochs_range = range(len(acc))
plt.figure(figsize=(12, 4))
plt.subplot(1, 2, 1)
plt.plot(epochs_range, acc, label='Training Accuracy')
plt.plot(epochs_range, val_acc, label='Validation Accuracy')
plt.legend(loc='lower right')
plt.title('Training and Validation Accuracy')
plt.subplot(1, 2, 2)
plt.plot(epochs_range, loss, label='Training Loss')
plt.plot(epochs_range, val_loss, label='Validation Loss')
plt.legend(loc='upper right')
plt.title('Training and Validation Loss')
plt.show()
from sklearn.metrics import confusion_matrix
import seaborn as sns
import pandas as pd
# 定义一个绘制混淆矩阵图的函数
def plot_cm(labels, predictions):
# 生成混淆矩阵
conf_numpy = confusion_matrix(labels, predictions)
# 将矩阵转化为 DataFrame
conf_df = pd.DataFrame(conf_numpy, index=class_names ,columns=class_names)
plt.figure(figsize=(8,7))
sns.heatmap(conf_df, annot=True, fmt="d", cmap="BuPu")
plt.title('混淆矩阵',fontsize=15)
plt.ylabel('真实值',fontsize=14)
plt.xlabel('预测值',fontsize=14)
val_pre = []
val_label = []
for images, labels in val_ds:#这里可以取部分验证数据（.take(1)）生成混淆矩阵
for image, label in zip(images, labels):
# 需要给图片增加一个维度
img_array = tf.expand_dims(image, 0)
# 使用模型预测图片中的人物
prediction = model.predict(img_array)
val_pre.append(class_names[np.argmax(prediction)])
val_label.append(class_names[label])
plot_cm(val_label, val_pre)
from sklearn import metrics
def test_accuracy_report(model):
print(metrics.classification_report(val_label, val_pre, target_names=class_names))
score = model.evaluate(val_ds, verbose=0)
print('Loss function: %s, accuracy:' % score[0], score[1])
test_accuracy_report(model)
1万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
