dental_caries_detection_v0_2.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "dental_caries_detection_v0.2.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "VTPyWg9YMkTJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zsPQIpBpN8PX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%tensorflow_version 1.x\n",
        "from keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Input\n",
        "from keras.preprocessing.image import load_img, img_to_array\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from itertools import chain\n",
        "import glob\n",
        "import copy\n",
        "import time"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "64bkKOAC5FuH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "'''fit the image to the background (when needed)'''\n",
        "import cv2\n",
        "\n",
        "def fit_to_background(image):\n",
        "  image = image.squeeze()\n",
        "  _, res = cv2.threshold(src=image, thresh=10, maxval=255, type=cv2.THRESH_BINARY)\n",
        "  res = np.where(res>0, 255, 0).astype(np.uint8)\n",
        "  _, cnts, _ = cv2.findContours(res, mode=cv2.RETR_EXTERNAL, method=cv2.CHAIN_APPROX_SIMPLE)\n",
        "  x, y, w, h = cv2.boundingRect(cnts[0])\n",
        "  image = image[y:y+h, x:x+w]\n",
        "  image = cv2.resize(src=image, dsize=(400, 400))\n",
        "  return image"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FjdcCou6Qg6u",
        "colab_type": "code",
        "outputId": "71ea3e58-cfef-400d-c3ae-56a6e08a969c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "path = './drive/My Drive/small_caries_dataset/Healthy/**.jpg'\n",
        "neg_images = list()\n",
        "for img_name in glob.glob(pathname=path):\n",
        "  img = load_img(path=img_name, color_mode='grayscale')\n",
        "  img = img_to_array(img=img, data_format='channels_last')\n",
        "  img = fit_to_background(img)\n",
        "  img = np.expand_dims(a=img, axis=-1)\n",
        "  neg_images.append(img)\n",
        "\n",
        "neg_images = np.array(neg_images)\n",
        "print('num healthy images:', len(neg_images))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "num healthy images: 238\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "da0a1497-76a3-4e5f-fdc4-6361f6d0e79f",
        "id": "UZyUcUD-VNdM",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "path = './drive/My Drive/small_caries_dataset/Caries/**.jpg'\n",
        "pos_images = list()\n",
        "for img_name in glob.glob(pathname=path):\n",
        "  img = load_img(path=img_name, color_mode='grayscale')\n",
        "  img = img_to_array(img=img, data_format='channels_last')\n",
        "  img = fit_to_background(img)\n",
        "  img = np.expand_dims(a=img, axis=-1)\n",
        "  pos_images.append(img)\n",
        "\n",
        "pos_images = np.array(pos_images)\n",
        "print('num carious images:', len(pos_images))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "num carious images: 45\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y5Ks_6s5ePrA",
        "colab_type": "code",
        "outputId": "f62cac3c-7219-4d53-a180-1d3a71fbedde",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "num_iter = int(len(neg_images) / len(pos_images))-1\n",
        "generated_pos_images = copy.deepcopy(x=pos_images)\n",
        "for _ in range(num_iter):\n",
        "  new_pos_images = copy.deepcopy(x=pos_images)\n",
        "  generated_pos_images = np.concatenate((generated_pos_images, new_pos_images))\n",
        "\n",
        "print('num repeated carious images:', len(generated_pos_images))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "num repeated carious images: 225\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ysz_jwbqXTMl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "neg_labels = [0 for _ in range(len(neg_images))]\n",
        "pos_labels = [1 for _ in range(len(generated_pos_images))]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cSoCU0GJXsSZ",
        "colab_type": "code",
        "outputId": "9ff17e30-930c-49f5-c273-913946330b34",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "X = np.concatenate((generated_pos_images, neg_images))\n",
        "y = np.array(pos_labels + neg_labels)\n",
        "\n",
        "print('num total dataset images:', len(X))\n",
        "print('num total dataset labels:', len(y))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "num total dataset images: 463\n",
            "num total dataset labels: 463\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bojZMZoUVWZ7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C5u6eHa9YCop",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras import Model\n",
        "inputA = Input(shape=X_train[0].shape)\n",
        "\n",
        "x = Conv2D(filters=64, kernel_size=3, strides=2, padding='same', activation='relu')(inputA)\n",
        "x1 = MaxPooling2D(pool_size=2, strides=2, padding='same')(x)\n",
        "x2 = Conv2D(filters=32, kernel_size=3, strides=2, padding='same', activation='relu')(x1)\n",
        "x3 = MaxPooling2D(pool_size=2, strides=2, padding='same')(x2)\n",
        "x4 = Flatten()(x3)\n",
        "x5 = Dense(units=10, activation='relu')(x4)\n",
        "x6 = Dense(units=1, activation='sigmoid')(x5)\n",
        "net = Model(inputs=inputA, outputs=x6)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PNntOAX_a1NX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras import optimizers\n",
        "\n",
        "net.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
        "net.summary()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "anFGnFI8mWn8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "\n",
        "aug = ImageDataGenerator(rescale=1./255, rotation_range=30, zoom_range=[1., 2.], \n",
        "                         width_shift_range=0.25, height_shift_range=0.25, \n",
        "                         shear_range=0.15, horizontal_flip=True, \n",
        "                         vertical_flip=True, fill_mode='nearest')\n",
        "\n",
        "test_aug = ImageDataGenerator(rescale=1./255)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LnvcWUNkbKXx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# from keras.callbacks import LambdaCallback\n",
        "# weights = LambdaCallback(on_epoch_end=lambda batch, logs: print(net.layers[10].get_weights()))\n",
        "\n",
        "t0 = time.time()\n",
        "net.fit_generator(generator=aug.flow(x=X_train, y=y_train, batch_size=8), \n",
        "                  steps_per_epoch=None, epochs=10, verbose=1)\n",
        "\n",
        "t1 = time.time()\n",
        "print('total training time: %.2f secs' % (round(t1-t0, ndigits=2)))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "In-QJohqmRD5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "y_pred = net.predict_generator(generator=test_aug.flow(x=X_test, batch_size=1),\n",
        "                               steps=len(X_test))\n",
        "y_test = np.reshape(a=y_test, newshape=(1, -1))\n",
        "\n",
        "print('num test-set images:', len(y_test[0]))\n",
        "print(y_test)\n",
        "y_pred = np.round(np.reshape(a=y_pred, newshape=(1, -1)))\n",
        "print(y_pred)\n",
        "# print(np.reshape(a=y_pred, newshape=(1, -1)))\n",
        "\n",
        "cm = confusion_matrix(y_true=y_test[0], y_pred=y_pred[0])\n",
        "print(cm)\n",
        "print('test-set acc:', (cm[0][0] + cm[1][1])/np.sum(cm))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RT-J9_bqWwho",
        "colab_type": "code",
        "outputId": "7b45158a-bac6-4061-80f0-eec7beb7a195",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 539
        }
      },
      "source": [
        "img = X_test[10] / 255.\n",
        "k_sz = 3\n",
        "prob_map = np.zeros(shape=img.shape)\n",
        "t0 = time.time()\n",
        "for row_idx in range(0, img.shape[1]-k_sz, k_sz):\n",
        "  for col_idx in range(0, img.shape[0]-k_sz, k_sz):\n",
        "    masked = copy.deepcopy(x=img)\n",
        "    masked[row_idx:row_idx+k_sz, col_idx:col_idx+k_sz] = 0\n",
        "    prob = net.predict(x=masked.reshape(1, img.shape[0], img.shape[1], 1))\n",
        "    # print(prob)\n",
        "    prob_map[row_idx:row_idx+k_sz, col_idx:col_idx+k_sz] = 1 - prob\n",
        "\n",
        "t1 = time.time()\n",
        "print('map generation time: %.2f secs' % (round(t1-t0, ndigits=2)))\n",
        "plt.imshow(img.squeeze(), cmap='gray')\n",
        "plt.show()\n",
        "plt.imshow(prob_map.squeeze(), cmap='jet')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "map generation time: 37.13 secs\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD8CAYAAAB3lxGOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD8CAYAAAB3lxGOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO2df6xc11Xvv+v53rrXrW9Tm9QyToQT\nYlQKlDQyaSoQ4rXqI80/LlKpUqQ2VJHCj1QCCSESkICiF6k8ARXVQ+UZNTRFQBN+qRYKlDQNQiA1\naSmpmx9Nc2lcNVeurcZtbqiNudfs98fZ27Nmzdo/zpk5c87MWR9pNGf22b/OzJx91l577bXIOQfD\nMIbL/+i6A4ZhdIsNAoYxcGwQMIyBY4OAYQwcGwQMY+DYIGAYA6e1QYCIbiaiZ4hog4juaqsdwzCm\ng9qwEyCiXQC+DOCtAJ4H8FkA73LOPTXzxgzDmIq2JIEbAWw4577inPsvAB8HcKyltgzDmIKVluo9\nBOBr7PPzAN4Yy7yXyF3ZUkeMxYP8uxOfOZr8SpF0A3gOBwGc/oZzk7daW4NAFiK6A8AdALAfwG91\n1RGjd6z59wvs847Isx0pd0FJN4B34w4A7/+qdq6tQWATwNXs81U+7TLOueMAjgPAtUQ2gA8EfqOu\n+vdtcSxZxWgQWFPOy7yynfAn5wOJ1s5QaWsQ+CyAI0R0Daqb/1YAP9VSW8YCsYrJm3Mb43/Edf++\nwz5zqSCUgci3ytLWWZ5Q35bIE/owdOmhlUHAObdDRO8D8EkAuwDc65x7so22DMOYjtZ0As65BwE8\n2Fb9Rn/hor2c3/MndGCNldkDYJ8/3vLvKywtsB/AC+x8YIelhTpDPVJSCG0/z46HKBWYxaBhDJzO\nVgeM/rKK0dNazqEhzoVjrsFfEe/hPFA90cPTdq9/P4PRk5k/9c/69wMANlj5UCag6QQuYCQBHFDa\nOY9J1jEpPQwBGwSMCc28dvOuAniNPw435AGMbhYu0nNxP5Q/599f/1pg7Uvj7ayz46t2A+tHfDtP\nVO+HXwWce7E6DoPAPgBP++M9/n0vRoPD9yp92M/a4fWEAeE61s8wcJzDaNqxrNh0wDAGjkkCxtgS\nWvgcRPbwxHwJo6djMPhYw7jIzRV58HWss2MA1ePYE57g5zCaGpy6CLz+VaP6AWB197jUAFRPZ82o\n6PI17QZwEWNss3Ze8u+HMa6UDFOQcN3nlHaWDRsEjIlpwA5LC+LxOoBTohyfN5/B5E2yqpR56l/G\nBxGguiF5Oyf/ZZQOAI+eHZ2/wN7Dcegrt0Y7eXF0foflewnjbLDyJ0Xe0MdwvKyDgE0HDGPgmCRg\nTFjh8TX2dXbukD8Oyjeu6eeKQf7ElKL0foxEbg5XKl63qzo+dWm8Dp4PmNxPwNf51zF66vM/uTQh\n5spEbs0o88s+LhMmCRjGwDFJYMmQTyu+lp8rwzffhPk+L3tO5JNIyzxeNhyfx6QOglsWAsCZS/F2\nQhmuj9BsGrYwWvqT9gOcc+x4G5Pzf033sGzYILBk8BsrfE79efkNwm8meQOuYty4Bxi/cdfYZ77j\nT9azvguAEPNXMS6my52CMbsFzUhoRUlL7VaU/dO2LMvvdNmw6YBhDByTBJYM+STbUY5XRPqWOM9N\ncrmijSsJgcmnf1Ac8qW6ID1cbuPS6BxXEPIlwlA+iPNfZeWDV4y9mLRL2IauqMxNjUKa9p3sgT6t\nWCapwCQBwxg4JgnMCb5UVmqB1mRJSjrs0BxpSOlAGu9o9Wxj/OkY8oS0QxjtKeD1SGXiKYwr40Ld\ngS2MlvZCPZuY/M64tMItFEO+ZzHpjWgLk/qGF1jfuNJyXSnDJallkgRsEJgx/Ibjfz5tDX0exDT5\nHO1PoCnYAlx85te1ruRbZXll3dp3wdM070Ecre4UXOHJpwhywANGUwDNxmDZsOmAYQycZR3cWiW1\n5MSVbnz9Xa7Fy3o05NKU3OdfVyRN+QiQ7aWevGEKcAHj6+pSwXYeo01H4cnKxfFYG2E6sJflk5IU\nt+nXrAj5EqJmr6C1fUE51upetptmqusholOofrNLAHacc0eJaB+A+1Ft0DoF4J3OuW9O103DMNpi\nFoPa/3TOfYN9vgvAw865D/gYhHcB+JUZtNNrNIOV2JO6dC6b8rWv1R2rT/7IJT+6lkc615B9kO3v\nweipvkfJp7XBjYW4BWOA6x1k+RLDHm0JVVv6k8rAVB2LThuSzTEAP+aP7wPwj1iCQYCL7nxPu+bT\nPhDWzTcwEm3PY3IawP+wsXZShHx877t2E2hWeXyPPReB5c0AjLz1hFWAG3YBX/Tr/ocwWvfnU4Tw\nHYT1/R/YBWxfGm9vG6NBgn+nfDvzXpEm+wZUAw73DXCeHQPj3z33lxBWGqRiM8ayrQ5Mqxh0AP6B\niP7VRxQCgAPOudP++OsYTQsNw+gh00oCP+Kc2ySi1wB4iIi+xE865xxFogvJMGR9h4/83EmlpnwK\ncMu5sDa+w8prS2RyzZyTs3jTfsyYAq0ULuGEp3no96lLoz6ew+gaeR83RdrmpUmlG1cqaoq4fYhP\nHcDOySU++dto0prciCS/Xy1t2ZhKEnDObfr3swD+BlU04jNEdBAA/Lu2fRzOuePOuaPOuaOlYliX\nSA1/OF5BfCTVymyzY22jDidX9zzYxuSNEPq1I85pN8wORvPuVYzXp13fCsu7EskT6k3BB9jc96yd\nT333y0bjQYCIXkFEe8MxgP8F4AkAJwDc5rPdBuAT03bSMIz2mGawOwDgb4go1PNnzrm/J6LPAniA\niG5HJUG+c/pu9oucJZvMt43xKUSpT/vS1YGSPsjj0na09PCU5ea/mxg3v5XIWAOA/j3sYHKatAN9\nJaBUGtL6I+uSLKs/QY3Gg4Bz7isAflBJfwHAW6bplGEY82Mo054oKeu/XJnY0yKlqJNt1qGJlWBJ\nu6VOM6TyjltH7sdoqW1FKaOFFA9Ld1wiWMHkMmfOOjBHqV0GX/pL/UbLpigc5CDAb6bUDRDLNy+l\nnKTrP19q8ON2DTxNhg07jMldgvz4AkbfNR9UpD2GNkXgHGB5Yx6MgCqqUrB72IPJFQfNhoO7X1sG\n56O2gcgwBs5gBgFti29Il8t9fIlKgz+Z5FKSdFFdV2pILT+mNv7E+lPSvuynPOZPQE2sD5z3r9CH\nIB2EQJ+r8JGBRF+3xbFsR5pklyzf8VDmvL7Qj9DXmBlybHlXLvEuw1LiYAYBwzB0Fn0Qi5Kaq/E5\n7VYkvbSMVDaV5otRuvQXa7PJcmFqyZLnkx6Bttj5LdY2z/es6NfnL47KBN3AtigjvRFp3oYBfbuv\n1lfNZXhQSnKrTv7dad+V3EotyywqSyUJlIrfTUa+mMWgli/Xn1mbojaZdqRI+Tco3YEY8u3B+A7C\nvdCvn1/DAYxvOJG7/FLfXWhbivlhmhemBiHtEManGHJKqFl28u9gn391pSyeBUs1CBiGUZ+lmg6k\nnhBNxOyY+JyaNpRY6JUuK3UV9KKpPYK2+UZGMjoLfWOPNu3QbAz4MqX29NV8GfLox1o9YPm0zU1a\nG6lNSYuGSQKGMXCWShJIMa21HWfWT+g+902D60ZSUlFsmTVlRbiNSc9CvB3NC5A2X+cGPWvsXXpR\nPrSr2t4czoc6g06Bb0nmrIn30v0gfWRpB4HSm6CuqfA24lrk0rqblNHylVg9lkxJUnEQtP5oN7EW\nQ3AH4x6DgPHIQYGgiAvtcU9J4bz8o65DD00e0q57BbDy7fHzr2HnD/nw5/uPAPBeMK7C+IpFaEeu\nLqwjHTth0bDpgGEMnKWVBEqZdgRvYjs+T3tzruzilLatKVQ1hZwmfvPyqU08Jdt2pU0/90XIJZlg\nKbj57ZENQFhuPMXOXwjhz58btXGG9SXUySMU8b6EtmUcR97HRWHR+tuYWd5wTTYdNSG3WzHnLQeY\n9KYb+lN3GgTo0wZtWqGZIPObS/7p+Dp8zriL18M3E4V6uOgudyMG82WwcysrAC6O6pe/GZ92cP8Q\nIW0PK1e6ualv2HTAMAbOIg1YU9FEA9+G1l7WH8vHn9axJ75U0E0j9ufgT1YN/l3xp3/dmAea0lWb\nXvC1es0nhLaiwAOWXjZNXtH9P2iSSeibpixcZEwSMIyBMxhJoMSSb1aklGDakhyfE/My51kZYHwd\nm7eTWhrUpJk6Ek4dS0kegxCo/lyyzDnoCkbejvQJqO0XyPX/DCbn8sCk4u/ZF0fnXsLkb8b7L5WB\nsu5UIJo+k5UEiOheIjpLRE+wtH1E9BARPevfX+3TiYg+REQbRHSSiG5os/N1KPEnIDeM8I0kfE/6\nfv/i6Txf4LB/rWK0KWYNk3vs5S65UEcoE/Jx192c0B9k6g7kbqDY+bBZhvc1rOHvYX3nm4ZCfwLX\nsTJ8Mw8vIzcQyeuRG3wC/FiLeCPdnm0DOPyKUdprRJvr/nrlNRwQx7Kt1LSpj5RMBz4K4GaRFuIN\nHgHwsP8MAG8DcMS/7gDw4dl00zCMtshOB5xz/0REh0VyLN7gMQAfc845AJ8hoiuI6CALS9YZMbGW\nK+JWlPO5JSuZj3+hZ9ixXM6Cck7WHcqnRHtgPFZhSmHF19r5cal7bbmx5wLGpwA7LD0gr2GD5dtS\n8kvfAvI8z5diA6PvRToyBUZSw2PfHtX/VUxueOLKxtDvM+w872tuObWvNFUMxuINHgLwNZbveYxi\nUo5BRHcQ0eeI6HOLbHdtGIvO1IrBVLzBTLnjAI4DwLWZ8k3s+3P5YoY+wPiTuYmzCGnJxqWMpnVK\nYgo/rS+alDJt2wGtbq4gK2kz6DvCsVwazW1U0ogFHJESGTdeKlW6yv4GNMXgPK1Dm9L0f3EmiPki\n3uAmgKtZvqsw7sGpEU002aVo1mZanTFNtra7TdMw873sWqhxrZ1YHnnM0+QPqrkek3li6bl+8I0/\n/JpjqyKxc7xtzU5AW6vPodUTu2G1aUwYRHI2Gk361jeaTgdi8QZPAHiPXyW4CcCLfdAHGIYRJzv4\nE9Gfo1ICfgcRPQ/gNwB8AHq8wQcB3IJKL3MewHtb6HOUUtEr9hRNrd9r7WxhtHQU9Br8C+Ui55pS\nJijNYu1o7ZZOifh6eDhOuVBvCrdhkFIRl0xCWmzJkovk2jRHXkMTqY8r+TQJLmdlmIMrlrvyCtWE\nktWBd0VOTcQb9KsCd07bqa5IBebMiYWxnXVAdePLvff8vLYDTfvDNiEnfgOTg2cTc2m+UYnXLUON\nxQZVPi2bh6OO0A9uAhycoIL1QZsG8gFPPgRCGTlNqrMKM2/MbNgwBs5gzIbbhivLctp4qQTjTwk+\nJdG2AJf2Y1bU0W7Lp6i2OpD7TrRtyk3gvwOndKoYnvB8OsClFfkbl2zyMknAMIxeslSSQFNlUUl5\nfq40AlFYG93GuOWctEqLlS+dE2tONwLTSgap7+cl6Aow3gdpZXhOqUd+lmW2lDw5Sp+60vov/J5h\nGXdHqeuMcpz7H/VVCgCWbBDoA1xU1LTKsR2DAc2MeVZGUjFynpJi50qmOykvQqX9iYn2kiYefUrq\n1qZvmu+EVD2l19AFNh0wjIFjksCUpERbvpmHLyVJ33cxa8O+W6DFnrzaPv5wfB76dc/C8q7Jn1la\nFqZ+zzqKRpne59/SJAHDGDgmCUxJkzlzygpRK19nLjkLw58mxOb/WlouyElTa7uuvfwuwmYhDRsE\nGlC6oqBp+negW89JZeI2RtZ2mnidu7lzf0h+PiWq5pRd2jEvK2/oab3upAa3tcx5DankK3EgKtsB\n8opB6RquT9h0wDAGjkkCLaItKWlLRdqyIX+S8W24/Mkq97TXUT6lpJlZ2hZI0b9OtKGUf4dt5Kdi\n8mmt5YvFleSWm/Im4W1rPg6bWCp2iUkChjFwTBJokTpfrvbUkvPnnB1+0x9z3gotudtQ6jrkLsKc\nIjV2PqWADEiJQUpTMSmt1O8hlxg03U4fMEmgRbjXGn48C6SXHL6RpdQ3Ab/5Z/nHlC7CNI8+Mv9K\n5JysM7aWz5WQPH9qY0+w3+B5pUtzfhxootxcE8d9cktug4BhDBybDrRIbgktEFPSaVZnmggdU26l\naFMkzV1DThGpef3RrAwh0rhjE16n9p3FXIKnwq/z7c51p1B93kBkkkAP4H84Hu2GR+WR7LBXqTgf\nE9ObTCGQOY7VzcVvHr2Ji8jhmF+XJpJLsZpHRzqMSdE+V4ZHVAr51lF5HNrL6jnA+nYII5/6vI+y\nnXVxzMOkd03TMGS/SUSbRPS4f93Czt3tw5A9Q0Q/3lbHDcOYDSXTgY8C+L8APibSP+ic+x2eQESv\nA3ArgO8D8J0APkVE3+OcuzSDvi4tMV8FMd/5gO7QVEYWCuc0cbh02hHrZ+pcqu7zSprMX2q3IPvI\nvzst8KmmmOURhLagi+3SipD7EzjLjjUbDl53KN+3sOZZScA590+YjAwV4xiAjzvnLjrnnkPldfjG\nKfpnGEbLTKMTeJ+PPHxviEoMC0M2V+SyYx39ACelE+BtlOoOSuvmfZ31MmVpH0rKrCZesaXN1JJn\nm9fahKaDwIcBfDeA6wGcBvC7dStwzh13zh11zh1dz2cfJCnbgiCyyxWIknV5ifxTyjpTMQHq1s3D\nq0+rqEx9P3KKxTdwaa7ltzHppi2sAmyLFy/Dr0OzC5HlgMlr75pGg4Bz7oxz7pJz7r8B/BFGIv8m\nWghDZhhGezQaBHz8wcBPAAgrBycA3EpEu4noGgBHADw2XReHS0qkDErAnCeiJlaKsXrk06vusiQv\npz0h5fkUwUKwRNLhy3WyDH+aa1Z9F6Bfg/Zd8Lrl8mSfaRqG7MeI6HoADsApAD8DAM65J4noAQBP\nofrf3GkrA4bRb5qGIftIIv89AO6ZplNDJrc9VvusLRFO02ZM0aVtgS55YpfA88nQXmuYvK5gZASM\nllL3Y7R8t45J1+7ade3BZAgxXmadlQvn9rIywdiILxtqT/490OMc9gEzG+4xpTd0yjw5FXNQ5tG8\n8vAydU1lU22mzmmitxY+XV7rmkhL2UzwHX2arwfte+ODUji/T8nHzZe1YLB9u+nMbNgwBk7fBqXB\nk7K2m2XdnFQ05lIPRJqVoaa05PliIdOl3YgWYUmTks6wdB7pqPQaYu3IMpusnQ2lH3xbs5Qi+HFf\nbAVMEjCMgWODwAIx7VJTnd2CWhnNKCZFTsyMLfHJHXja0p3mlEPmK7ne3PmcU5F9/qXpEDjabtC+\nLB3adGCBmFZ8nFX52OqBhIv+UiFX0o72mU8r5EDA6865AI+1JdGmA7Fpwx7lfCqGgk0HDMPoBSYJ\n9JgmS3KzRtq8A7pCT1t+gzifQlsCTBHEdNnH0vM5NIcp2hLqXpaPO0iVon6wPuT1aVJGF9gg0DPa\nuPG19XKwNHlTX1DyauvcsTltan4ccwHGIy+ljGhCmTVM+lvg7e5DfPUh9COc4zenXFGIRQ6SfVzB\nSOsvzZKByuiIH2t1dIVNBwxj4Jgk0DNSDjJzZrryCR/SUubFXCTVrO1S7eXMlGObm7ZF2opIk/19\nCeNSSkCTKMLxOZY3FU0ohmbaG/rDxf2z7Nwedj5Vpg9TAI5JAoYxcEwS6BmlTwvNvl8jNR+OpfN9\nAql6tD6F8rJM6bIi70dK0agp7krr5+dWkY4mpNWt+XLk0oxWZhXjew/6hA0CAyK2uaZEPOXr8m1s\nINJuHE0pqXlN4mI/P9bc1kkxv4nBTm7g0NI0ZWFfsOmAYQwckwR6RpP995pSjVOqvAvERG3Zt9TG\nI61vAW0DUeoatLSU/QJQKRO1aEIpa0QOl1Jk2zwCEbcYXGPHsh5NKumLRGCSgGEMHJMEegZfzpK2\n6CVbgeVcV7O75/PldYw8wR5gZWR5Pt8OVnIxQxventw7wPPVsQ7USDkj2aec1xSVB1D5xwPGvRFx\nn4Tcwi+cC9d9yL/LJUBZ5ioAz7I2gWrrcx8Mhkp8DF6NKvrQAVQ+BY87536fiPYBuB9V2LdTAN7p\nnPsmERGA3wdwCyqjrp92zn2+ne4vH/yGnkU8hpjiS9sVx+0SSm5QuZGmtIy0OIwpBuusKIR6Aqci\n5+Xgdkacl8fB9Xisra/69x2MLBgvYNLOgrdzjpXpg+1AyXRgB8AvOedeB+AmAHf6cGN3AXjYOXcE\nwMP+MwC8DZWX4SMA7kAVo8AwjJ5SEobsdHiSO+deAvA0KinoGID7fLb7ALzdHx8D8DFX8RkAVwgX\n5UYhpS61U8jpQMlWXmndt418dKNcO6mgH/K4rpvu0uuKrd+XlKuTbxWTLse1FzCb33haaikGiegw\ngDcAeBTAAefcaX/q6xhNdQ6hIBSZhSEzjH5QPAgR0SsB/BWAX3TObVVT/wrnnCMiV6dh59xxAMcB\n4NqaZYeC9jRNWdFpii/pTbcEbtFWYpkny6Tm8rHlumnmxrH25Lw8dy2aErXErXpoL5dP83ko+9OF\nbqDof0FEq6gGgD91zv21Tz5DRAedc6e9uB/2UmzCQpG1gmauCpaWW/OvY4Mg17z5SoAGXzvndgCy\nnpgtQ2nfUtfPvx/en5TbdGmNKOu/oJThdfJBWcZL4GjKT/49d7lKkJ0OeG3/RwA87Zz7PXbqBIDb\n/PFtAD7B0t9DFTcBeJFNGwzD6BklksAPA3g3gC8S0eM+7VcBfADAA0R0O6qVknf6cw+iWh7cQLVq\n8t6Z9nhAaP7pcptiUlZ8mojL17w1ewPNJXdqo5FsR/aHr7trxETl1HZnbRqkRS3SNl3JyEDyiczr\n4U9/uU1Z9kHaCezDaNkyBCzJbYueF9lBwDn3zwAocvotSn4H4M4p+2VA3+yS8ydQOv9P6RS4nYD8\ns8s6VpVjrU7NhHdWaNMBbQDi7ZfudOTEphC8D7Ey6yxN002U/sZtYGbDhjFwul6iNASa+K2hicja\n0zj25OUOMsParlSkAeOx9vgxb1cep+DKwhKbg1Q+SYm1pfw+zrHjrUjb8nvh7TzP8mkWg6HMKVbm\nBVamD9GITBIwjIFjksACoymTtC23OeUdVxJKuCKtriMO7emW88CjLZs1mSfHysg06U04JUlp/Q6b\nvHIWi3zTlrb1mbc3b6nABoEFJrVLLlDieit1U2sbjTRiUxGpGNQiB2l1NUVbUUkhb0R5o8cGUZk/\ntumKb0QqoYtpgU0HDGPgmCSwwKR8BpQuEeamA9NQakY7S3LXn1oi5GgSRc7smIv5sl6+aUjzk2iK\nQcMwOsMkgR6TWyqT/vtKPQdLwnx1PZlrRKklo/bE4+Vy+oqcYVSqLLf0KzUC0uqPGUCF74xb/3FF\npmYpmbJ61NAsHNuQGGwQWFBSf1xA38QD5TyU89pGG21PfWznYC4gaWoFQ7uuWGzEVDvaBqItlA90\nWt94G9JmYBvjG4jkxiktnDmQHvDmNUWw6YBhDByTBJaUWBzAgNwKW0JKmTbNsl6OWXlX4k90zWlo\nkzo1v4SlHoxi+xs02lQgmiRgGAPHJIElolQBxp9+2tbemDIs9xRKeT2SeZrC20gZ9mhbgLV+aU9y\neSzTuJUh1ztokZybLEXmmLWS0AaBJSDnmluL6KNp7rWbOLdVONUHjRLPQvLPrblVy61GxDwCaXXz\n8qkbSxtMwqYhbu6s2RbwDU0pN2MxpPPXWWLTAcMYOCYJDICc8kmLWiQ3uciIv0D5ejdHW66Um3Sk\nr769GJ++hPakSM6XQ7XpAD/misFSx5/adCD0VbMLCH0CKnuCcMwjS6WkIu07NzsBI4v8QUt+4PBH\n1jwL8ZtWu/m1dnJGOrlNOlJbr01PeD2yf7IM38Qjb9YLSr7UsSzD/QGkBkU+HTiv5EutYGh9mCUl\njkavJqJHiOgpInqSiH7Bp/8mEW0S0eP+dQsrczcRbRDRM0T04y323zCMKSl5UIQwZJ8nor0A/pWI\nHvLnPuic+x2e2YcouxXA9wH4TgCfIqLvcc5dmmXHjfnAn1Cl/gs5Kakgtkmn6Zq4tDws2dpcsuqR\n0sanfAPw8zH6sDowTRiyGMcAfNw5d9E59xwqr8M3zqKzhmHMnmnCkAHA+4joJBHdS0Sv9mmHUBCG\nzOgP2/4VYvqtimO+Fh57sjWxGORt5Aht876m+iPn1jvKMVDW79CmBt84xI+3xYvD+yCjHpc83WWd\n03pwLh4EZBgyVNGGvxvA9QBOA/jdOg1bLML+EG6mFYzEY35c8sdsomHmbeQI/eB9Td0M/AbU2im9\nrhyhbqBSTAbl5Kp48bRYGXkNEOncEGqWvh+KBgEtDJlz7oxz7pJz7r8B/BFGIv8mCsKQOeeOO+eO\nOueO1t3ZZRjG7GgchkyEG/8JAE/44xMAbiWi3UR0DYAjAB6bXZeNeZALJZ6CP+n4cahDa4PXLcX8\nNSVNPmV5W7EoR1p5+RTWXuHJq7W5B6N1fw0uweSmBinWMW6TMMsNWyVS3A9DD0P2LiK6HoBD5Vb9\nZwDAOfckET0A4Cnf1zttZcAw+ss0YcgeTJS5B8A9U/TL6BC+lBaYpVVZ6inGLQZjG5lCHZofP76M\nmZJapMVfKCOXNDUHInU2U9XVO/C6uWWithdEu9Ymy4dmMWgk/wSlUYZiG4NKkF6HXlDylJr2anWm\n/AVwpTQ3RU61q20gSuWR7aRWA2L93ozkydVXgm0gMoyBY4OAMcEsxcOcbUHM65FU+GnLa5JSqaVO\nGa3NFXGeTxdiU50ma/nSD2RsWVBTeNbBBgHDGDimEzAmaDK/z+0YzJXJLXmVzHXrWgKWlCn1NlSa\nXofSXYS5PuawQcCYCaUDR4lzzZKBQ6tD+vavqzHXYgPELBLDseYbQNu8xNE2OeXMgHMblbS6S7Hp\ngGEMHJMEjCSlUXxK/0jS+YZWXj7NuPSQcp/O2ZPJW7qsGHui58pLYrYFqbSY16NZOxixQcCYuDGa\naNmbUCdMWM4bUahvVSmTq1sj1U5MjE8ZCOVWDbTpQGmZnIPUHDYdMIyBY5KA0ar/uhR13JRLNC9B\nwLgozfPyc7KebeW89oTPhXNPrS6Urt2vID8VmTUmCRjGwLFBwJigzbiCJaQs9HLlFgGur5C6i1T0\n5rZYlO/NmCNd/ylSLscDMeVbKl6AVq6uEVKqvpxtgUyLKRgDpRuDpjUWMknAMAaODQJGMZp3n7pI\nM135Asa96ABV9B65gUjzJ+1HtOEAAAkVSURBVMj9Dpb2LZePexYK8L6VehbiT2jNIWmub6VOVUsd\nw3JsEDCMgdP19M9YYJouLWpx+1JOQC6wtlIGORcixylkPEBpwJSbb3OnIjlnIam+59rTohprn5vo\nB0ocjb6ciB4joi/4MGTv9+nXENGjPtzY/UT0Mp++23/e8OcPF/bFWDCarmOnfAIE7z4lf2DNrfes\n0OrkU5LU+TZoIuaXUjIduAjgzc65H0QVY+BmIroJwG+jCkN2HYBvArjd578dwDd9+gd9PsMwekpJ\nGDLnnPsP/zEMQg7AmwH8pU+/D8Db/fEx/xn+/Fu823LDADBSAmoSgfaU1dyUcxfedZWBqX7xNjWx\nexvj0krMbXqbzLqd0uAju7y78bMAHgLw7wC+5ZwL189DjR2CD0Pmz78IYP8M+2x0xKz+fOHmj7nj\nSrXDownJ6cC0fZNxC0oGFT6QtTkd4HQxHYCPNHQ9qmhCNwJ47bQNWxgyw+gHtZYInXPfAvAIgDcB\nuIKIwiDIQ41twoch8+dfBcWLtIUhW3xm9eRtguYAlK/l5+rVpBBprxAj5UotJcEEpM1DW0t0M7MT\nIKIriegKf7wG4K2owpM/AuAdPtttAD7hj0/4z/DnP+2cc+VdNwxjnpQMQgcB3EdEu1ANGg845/6W\niJ4C8HEi+t8A/g1VvEL49z8hog0A5wDc2kK/jR7D7eHD0zmsoZf6IQzlw2dt3V863yjxXxhI+SqU\nbcg1fN6fUt9/pT4Yeb5YmZQE1koEIufcSQBvUNK/glEkYp7+nwB+skYfjCVDmshyYgq8WNBSWcdO\nJE/Ip3lHKoUbLcV8D8j+aH2Txk9aO3XPld7UrRgLGYax3JjZsNE76rgIB8p9FeaoM5UofcqWTjt4\nH7TraHKNpd+jSQKGMXBMEjDmyjQegCXa/H0anUApXXheavN6bBAwDA9X6OVE6XA+tTMQqD9gdHFD\n2nTAMAaOSQJG72niL1B7AueceMYCgErl3hpG+/ubRGbKRUGalWLQlggNwyjCJAGj9zQJla5JBSVG\nPDkJAqgMhDSb/CZu0ptY+M0aGwSMpWfaP3kq0lHMNdmsbBfmgU0HDGPg2CBgLCWaa/M6ZTS4a/O6\ndc8rrmATFkFaMQZO081AuePSdqTtgLbKIAcHWVddM+N5YpKAYQwcGwSMpSQ2HUiJ77H0lAejJv7+\nUgFJZbuxPswSGwQMY+CYTsAwGFp0JOkYJeU0JVVvaltx7kZs057ABgHDYNS9IXIbiHJlSzcYtWlM\nNE0Yso8S0XNE9Lh/Xe/TiYg+5MOQnSSiG1rsv2EYU1Iy8IUwZP9BRKsA/pmI/s6f+2Xn3F+K/G8D\ncMS/3gjgw/7dMDqhztO9rqXfNL4FuvBLoDFNGLIYxwB8zJf7DKr4BAen76phGG3QKAyZc+5Rf+oe\nL/J/kIh2+7TLYcg8PESZYcyFJlZ9ody8FGV9Ucg1CkNGRN8P4G5U4ch+CMA+AL9Sp2ELQ2bMC35j\np27yujv/6vYhlt71YNA0DNnNzrnTXuS/COCPMYpBcDkMmYeHKON1WRgyw+gBTcOQfSnM833Y8bcD\neMIXOQHgPX6V4CYALzrnTrfSe8OoSZ2pQSBlFTjrJ/lq5LhNpglD9mkiuhIAAXgcwM/6/A8CuAXA\nBoDzAN47+24bhjErpglD9uZIfgfgzum7ZhizR/7hS56CdX0Z1u1DrO55eRuyvQOGkUFOB+Ylps+r\nHRsEDGPgdL06YRi9JzUdmGe7bWGSgGEMHBsEDGPg2HTAMGqyivqi+rT+BNrEJAHDGDg2CBhGTWSc\nwqZP8b6I4TYIGMYC0MShaSk2CBjGwOmLRGIYCwNXDDYxGy5VKjZRQDbBJAHDGDgmCRhGTaZ9Ooe5\nfc7b8LwsBm0QMIwpaHKjamHNu8SmA4YxcPoyGBnGYOBLfaU3YKfBRwzDWG5MEjCMOSAjDPcJkwQM\nY840cXbaJjYIGMbA6ZtkYhhLT99uul705zkcxLtxR9fdMIxBQpWH8I47QfQSgGe67kdLfAeAb3Td\niRZY1usClvfavss5d6VM7IUkAOAZ59zRrjvRBkT0uWW8tmW9LmC5r03DFIOGMXBsEDCMgdOXQeB4\n1x1okWW9tmW9LmC5r22CXigGDcPojr5IAoZhdETngwAR3UxEzxDRBhHd1XV/6kJE9xLRWSJ6gqXt\nI6KHiOhZ//5qn05E9CF/rSeJ6Ibuep6GiK4mokeI6CkiepKIfsGnL/S1EdHLiegxIvqCv673+/Rr\niOhR3//7iehlPn23/7zhzx/usv+t4Jzr7AVgF4B/B3AtgJcB+AKA13XZpwbX8KMAbgDwBEv7PwDu\n8sd3Afhtf3wLgL8DQABuAvBo1/1PXNdBADf4470AvgzgdYt+bb5/r/THqwAe9f19AMCtPv0PAfyc\nP/55AH/oj28FcH/X1zDz76TjH+RNAD7JPt8N4O6uv5QG13FYDALPADjojw+isoMAgP8H4F1avr6/\nAHwCwFuX6doA7AHweQBvRGUctOLTL/8vAXwSwJv88YrPR133fZavrqcDhwB8jX1+3qctOgecc6f9\n8dcBHPDHC3m9XgR+A6qn5sJfGxHtIqLHAZwF8BAqafRbzrmwuY/3/fJ1+fMvAtg/3x63S9eDwNLj\nqkfIwi7BENErAfwVgF90zm3xc4t6bc65S8656wFcBeBGAK/tuEud0vUgsAngavb5Kp+26JwhooMA\n4N/P+vSFul4iWkU1APypc+6vffJSXBsAOOe+BeARVOL/FUQUzOh53y9flz//KgAvzLmrrdL1IPBZ\nAEe8ZvZlqBQvJzru0yw4AeA2f3wbqvl0SH+P16TfBOBFJlr3CiIiAB8B8LRz7vfYqYW+NiK6koiu\n8MdrqPQcT6MaDN7hs8nrCtf7DgCf9hLQ8tC1UgKVVvnLqOZlv9Z1fxr0/88BnEblC/J5ALejmjM+\nDOBZAJ8CsM/nJQB/4K/1iwCOdt3/xHX9CCpR/ySAx/3rlkW/NgCvB/Bv/rqeAPDrPv1aAI8B2ADw\nFwB2+/SX+88b/vy1XV/DrF9mMWgYA6fr6YBhGB1jg4BhDBwbBAxj4NggYBgDxwYBwxg4NggYxsCx\nQcAwBo4NAoYxcP4/2EhmD3N5yJYAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}