2DCNN_specg.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Dropout, Flatten\n",
    "from tensorflow.keras.layers import Conv2D, MaxPooling2D\n",
    "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "from keras.utils import to_categorical\n",
    "import matplotlib.pyplot as plt\n",
    "import Eval_model\n",
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cnn_model_2d(x_train, y_train, x_test, y_test, epochs =50):\n",
    "    nClass = 4\n",
    "    model = Sequential()\n",
    "    model.add(Conv2D(32, (3,3), activation = 'relu', input_shape = x_train.shape[1:]))\n",
    "    model.add(Conv2D(64, (3,3), activation = 'relu'))\n",
    "    model.add(MaxPooling2D(pool_size = (2,2)))\n",
    "    model.add(Dropout(0.25))\n",
    "    model.add(Conv2D(64, (3,3), activation = 'relu'))\n",
    "    model.add(Conv2D(64, (3,3), activation = 'relu'))\n",
    "    model.add(MaxPooling2D(pool_size = (2,2)))\n",
    "    model.add(Dropout(0.25))\n",
    "    \n",
    "    model.add(Flatten())\n",
    "    model.add(Dropout(0.2))\n",
    "    model.add(Dense(512, activation = 'relu'))\n",
    "    model.add(Dense(128, activation = 'relu'))\n",
    "    #model.add(Dropout(0.2))\n",
    "    model.add(Dense(nClass, activation = 'softmax' ))\n",
    "\n",
    "    model.compile(optimizer = 'adam', loss = 'categorical_crossentropy', \n",
    "                  metrics = ['accuracy'])\n",
    "    callbacks = [EarlyStopping(monitor='val_loss', patience = 20), \n",
    "                 ModelCheckpoint(filepath='best_model.h5', monitor='val_loss', save_best_only=True)\n",
    "                 ]\n",
    "    history = model.fit(x_train, y_train, batch_size = 32, epochs = epochs,\n",
    "                        callbacks = callbacks, validation_data = (x_test, y_test) )\n",
    "    \n",
    "    #model.load_weights('best_model.h5')\n",
    "    model.summary()\n",
    "    return model, history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def InputPreprocess(x_train, y_train, x_test, y_test, IMG_SIZE = 64):\n",
    "    \n",
    "    #reshape\n",
    "    x_train /= 255\n",
    "    x_test  /= 255\n",
    "    \n",
    "    x_train = np.reshape(x_train, (-1, IMG_SIZE, IMG_SIZE, 1))\n",
    "    x_test = np.reshape(x_test, (-1, IMG_SIZE, IMG_SIZE, 1))\n",
    "\n",
    "    \n",
    "    y_true = y_test\n",
    "    #to categorical label\n",
    "\n",
    "    y_train = to_categorical(y_train, num_classes = 4, dtype = 'int8')\n",
    "    y_test = to_categorical(y_test, num_classes = 4, dtype = 'int8' )\n",
    "    \n",
    "    return x_train, y_train, x_test, y_test, y_true"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pickle.load(open('train_specg_input.pk1', 'rb'))\n",
    "test = pickle.load(open('test_specg_input.pk1', 'rb'))\n",
    "x_train = []\n",
    "y_train = []\n",
    "\n",
    "for lb, sg in train:\n",
    "    x_train.append(sg)\n",
    "    y_train.append(lb)\n",
    "\n",
    "x_train = np.array(x_train).astype('float32')\n",
    "y_train = np.array(y_train).astype('int8')\n",
    "\n",
    "x_test = []\n",
    "y_test = []\n",
    "for lb, sg in test:\n",
    "    x_test.append(sg)\n",
    "    y_test.append(lb)\n",
    "\n",
    "x_test = np.array(x_test).astype('float32')\n",
    "y_test = np.array(y_test).astype('int8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, y_train, x_test, y_test, y_true = InputPreprocess(x_train, y_train, x_test, y_test, 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 8528 samples, validate on 300 samples\n",
      "Epoch 1/50\n",
      "8528/8528 [==============================] - 23s 3ms/sample - loss: 0.9811 - accuracy: 0.5876 - val_loss: 1.2760 - val_accuracy: 0.5000\n",
      "Epoch 2/50\n",
      "8528/8528 [==============================] - 21s 2ms/sample - loss: 0.9107 - accuracy: 0.6060 - val_loss: 1.1630 - val_accuracy: 0.5467\n",
      "Epoch 3/50\n",
      "8528/8528 [==============================] - 21s 3ms/sample - loss: 0.8883 - accuracy: 0.6194 - val_loss: 1.1617 - val_accuracy: 0.5233\n",
      "Epoch 4/50\n",
      "8528/8528 [==============================] - 21s 3ms/sample - loss: 0.8754 - accuracy: 0.6290 - val_loss: 1.0125 - val_accuracy: 0.5533\n",
      "Epoch 5/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.8370 - accuracy: 0.6508 - val_loss: 0.9578 - val_accuracy: 0.5633\n",
      "Epoch 6/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.7895 - accuracy: 0.6664 - val_loss: 0.9226 - val_accuracy: 0.5767\n",
      "Epoch 7/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.7504 - accuracy: 0.6849 - val_loss: 0.9375 - val_accuracy: 0.5833\n",
      "Epoch 8/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.7024 - accuracy: 0.7045 - val_loss: 0.7628 - val_accuracy: 0.6700\n",
      "Epoch 9/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.6445 - accuracy: 0.7314 - val_loss: 0.6806 - val_accuracy: 0.7200\n",
      "Epoch 10/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.5782 - accuracy: 0.7595 - val_loss: 0.6267 - val_accuracy: 0.7500\n",
      "Epoch 11/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.5025 - accuracy: 0.7905 - val_loss: 0.5920 - val_accuracy: 0.7833\n",
      "Epoch 12/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.4005 - accuracy: 0.8347 - val_loss: 0.5476 - val_accuracy: 0.8300\n",
      "Epoch 13/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.3306 - accuracy: 0.8610 - val_loss: 0.4781 - val_accuracy: 0.8900\n",
      "Epoch 14/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.2587 - accuracy: 0.8935 - val_loss: 0.4698 - val_accuracy: 0.9067\n",
      "Epoch 15/50\n",
      "8528/8528 [==============================] - 22s 3ms/sample - loss: 0.2140 - accuracy: 0.9173 - val_loss: 0.4644 - val_accuracy: 0.9300\n",
      "Epoch 16/50\n",
      "8528/8528 [==============================] - 21s 3ms/sample - loss: 0.1693 - accuracy: 0.9381 - val_loss: 0.4504 - val_accuracy: 0.9467\n",
      "Epoch 17/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.1344 - accuracy: 0.9481 - val_loss: 0.5091 - val_accuracy: 0.9467\n",
      "Epoch 18/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.1013 - accuracy: 0.9633 - val_loss: 0.5410 - val_accuracy: 0.9467\n",
      "Epoch 19/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0946 - accuracy: 0.9661 - val_loss: 0.6847 - val_accuracy: 0.9467\n",
      "Epoch 20/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0695 - accuracy: 0.9737 - val_loss: 0.6258 - val_accuracy: 0.9500\n",
      "Epoch 21/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0766 - accuracy: 0.9764 - val_loss: 0.5918 - val_accuracy: 0.9433\n",
      "Epoch 22/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0658 - accuracy: 0.9778 - val_loss: 0.6103 - val_accuracy: 0.9500\n",
      "Epoch 23/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0450 - accuracy: 0.9845 - val_loss: 0.7085 - val_accuracy: 0.9433\n",
      "Epoch 24/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0552 - accuracy: 0.9830 - val_loss: 0.6325 - val_accuracy: 0.9433\n",
      "Epoch 25/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0426 - accuracy: 0.9863 - val_loss: 0.7702 - val_accuracy: 0.9467\n",
      "Epoch 26/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0392 - accuracy: 0.9870 - val_loss: 0.8897 - val_accuracy: 0.9467\n",
      "Epoch 27/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0460 - accuracy: 0.9850 - val_loss: 0.7324 - val_accuracy: 0.9500\n",
      "Epoch 28/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0411 - accuracy: 0.9859 - val_loss: 0.8754 - val_accuracy: 0.9500\n",
      "Epoch 29/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0451 - accuracy: 0.9849 - val_loss: 0.7387 - val_accuracy: 0.9500\n",
      "Epoch 30/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0346 - accuracy: 0.9900 - val_loss: 0.9357 - val_accuracy: 0.9500\n",
      "Epoch 31/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0454 - accuracy: 0.9849 - val_loss: 0.8383 - val_accuracy: 0.9533\n",
      "Epoch 32/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0312 - accuracy: 0.9883 - val_loss: 0.7657 - val_accuracy: 0.9533\n",
      "Epoch 33/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0411 - accuracy: 0.9865 - val_loss: 0.8049 - val_accuracy: 0.9500\n",
      "Epoch 34/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0391 - accuracy: 0.9876 - val_loss: 0.8087 - val_accuracy: 0.9533\n",
      "Epoch 35/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0313 - accuracy: 0.9905 - val_loss: 0.8422 - val_accuracy: 0.9533\n",
      "Epoch 36/50\n",
      "8528/8528 [==============================] - 20s 2ms/sample - loss: 0.0390 - accuracy: 0.9882 - val_loss: 1.0046 - val_accuracy: 0.9533\n",
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_4 (Conv2D)            (None, 98, 98, 32)        320       \n",
      "_________________________________________________________________\n",
      "conv2d_5 (Conv2D)            (None, 96, 96, 64)        18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 48, 48, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 48, 48, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_6 (Conv2D)            (None, 46, 46, 64)        36928     \n",
      "_________________________________________________________________\n",
      "conv2d_7 (Conv2D)            (None, 44, 44, 64)        36928     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 22, 22, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 22, 22, 64)        0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 30976)             0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 512)               15860224  \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 128)               65664     \n",
      "_________________________________________________________________\n",
      "dropout_5 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 4)                 516       \n",
      "=================================================================\n",
      "Total params: 16,019,076\n",
      "Trainable params: 16,019,076\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model, history = cnn_model_2d(x_train, y_train, x_test, y_test, 50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_4 (Conv2D)            (None, 98, 98, 32)        320       \n",
      "_________________________________________________________________\n",
      "conv2d_5 (Conv2D)            (None, 96, 96, 64)        18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 48, 48, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 48, 48, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_6 (Conv2D)            (None, 46, 46, 64)        36928     \n",
      "_________________________________________________________________\n",
      "conv2d_7 (Conv2D)            (None, 44, 44, 64)        36928     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 22, 22, 64)        0         \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 22, 22, 64)        0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 30976)             0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 512)               15860224  \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 128)               65664     \n",
      "_________________________________________________________________\n",
      "dropout_5 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 4)                 516       \n",
      "=================================================================\n",
      "Total params: 16,019,076\n",
      "Trainable params: 16,019,076\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 95.33%\n",
      "F1 score = 93.52\n",
      "Sensitivity: Normal = 0.967, AFib = 0.940, Others = 0.914, Noise = 1.000\n",
      "Specificity: Normal = 0.987, AFib = 0.996, Others = 0.996, Noise = 0.963\n",
      "Positive predictive value: Normal = 0.986, AFib = 0.979, Others = 0.985, Noise = 0.750\n",
      "Negative predictive value: Normal = 0.967, AFib = 0.988, Others = 0.974, Noise = 1.000\n",
      "Confusion matrix, without normalization\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Normalized confusion matrix\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "Eval_model.get_summary(x_test, y_true, model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 95.33%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "Eval_model.Plot_Acc_and_Loss(history, x_test, y_test, model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}