U_net_tumor_.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Increase RAM Reference Notes By Techhawa .ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "y1bbp2IY8HTp"
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import glob\n",
        "import keras as k\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "import glob\n",
        "from keras import Input\n",
        "from keras.layers import Lambda,LeakyReLU,Conv2D,Dense,Dropout,Conv2DTranspose,MaxPooling2D,concatenate,UpSampling2D,BatchNormalization\n",
        "from keras.models import Model\n",
        "import cv2\n",
        "import PIL\n",
        "import matplotlib.pyplot as plt\n",
        "import skimage.io as io\n",
        "import pandas as pd\n",
        "from keras.preprocessing.image import array_to_img,load_img,img_to_array\n",
        "from skimage import io\n",
        "import os\n",
        "from skimage.color import rgb2gray\n",
        "from scipy import ndimage\n",
        "!pip install MedPy\n",
        "from medpy.io import load"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FzXbmWh38W7s"
      },
      "source": [
        "height=240\n",
        "width=240\n",
        "channle=1\n",
        "\n",
        "inputs=Input((height,width,channle))\n",
        "\n",
        "c1=Conv2D(64,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(inputs)\n",
        "c1=BatchNormalization()(c1)\n",
        "c1=Conv2D(64,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(c1)\n",
        "p1=MaxPooling2D()(c1)\n",
        "\n",
        "c2=Conv2D(128,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(p1)\n",
        "# c2=BatchNormalization()(c2)\n",
        "c2=Conv2D(128,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(c2)\n",
        "p2=MaxPooling2D()(c2)\n",
        "\n",
        "c3=Conv2D(256,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(p2)\n",
        "# c3=BatchNormalization()(c3)\n",
        "c3=Conv2D(256,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(c3)\n",
        "p3=MaxPooling2D()(c3)\n",
        "\n",
        "c4=Conv2D(512,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(p3)\n",
        "# c4=BatchNormalization()(c4)\n",
        "c4=Conv2D(512,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(c4)\n",
        "drop4=Dropout(.5)(c4)\n",
        "p4=MaxPooling2D()(drop4)\n",
        "\n",
        "c5=Conv2D(1024,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(p4)\n",
        "# c5=BatchNormalization()(c5)\n",
        "c5=Conv2D(1024,3,padding=\"same\",activation=\"relu\",kernel_initializer=\"he_normal\")(c5)\n",
        "drop5=Dropout(.5)(c5)\n",
        "\n",
        "u6=Conv2D(512,2,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(UpSampling2D()(drop5))\n",
        "merge6=concatenate([u6,drop4])\n",
        "c6=Conv2D(512,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(merge6)\n",
        "c6=BatchNormalization()(c6)\n",
        "c6=Conv2D(512,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(c6)\n",
        "\n",
        "u7=Conv2D(256,2,activation=\"relu\",kernel_initializer=\"he_normal\",padding=\"same\")(UpSampling2D()(c6))\n",
        "merge7=concatenate([u7,c3])\n",
        "c7=Conv2D(256,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(merge7)\n",
        "# c7=BatchNormalization()(c7)\n",
        "c7=Conv2D(256,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(c7)\n",
        "\n",
        "u8=Conv2D(128,2,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(UpSampling2D()(c7))\n",
        "merge8=concatenate([c2,u8])\n",
        "c8=Conv2D(128,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(merge8)\n",
        "# c8=BatchNormalization()(c8)\n",
        "c8=Conv2D(128,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(c8)\n",
        "\n",
        "u9=Conv2D(64,2,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(UpSampling2D()(c8))\n",
        "merge9=concatenate([u9,c1])\n",
        "c9=Conv2D(64,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(merge9)\n",
        "# c9=BatchNormalization()(c9)\n",
        "c9=Conv2D(64,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(c9)\n",
        "\n",
        "c9=Conv2D(64,3,activation=\"relu\",padding=\"same\",kernel_initializer=\"he_normal\")(c9)\n",
        "output=Conv2D(1,1,activation=\"sigmoid\")(c9)\n",
        "\n",
        "model=Model(inputs,output)\n"
      ],
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ilFxXcuk_GR1"
      },
      "source": [
        "model.compile(optimizer=\"adam\",loss=\"binary_crossentropy\",metrics=[\"acc\"])"
      ],
      "execution_count": 40,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T6G_thgD8ne8"
      },
      "source": [
        "# !unzip \"/content/drive/My Drive/brain-tumor-segmentation-in-mri-brats-2015.zip\"\n",
        "\n",
        "images=[]\n",
        "labels=[]\n",
        "\n",
        "path=\"/content/train/HGG/\"\n",
        "list_file=os.listdir(path)\n",
        "for file in list_file:\n",
        "  list_=os.listdir(path+file+\"/\")\n",
        "  list_=sorted(list_)\n",
        "  image,_=load(path+file+\"/\"+list_[0])\n",
        "  label,_=load(path+file+\"/\"+list_[4])\n",
        "  for i in range(image.shape[2]):\n",
        "    images.append(image[:,:,i])\n",
        "    labels.append(label[:,:,i])\n"
      ],
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4jKvtyfeT2iJ"
      },
      "source": [
        "path=\"/content/test/HGG_LGG/\"\n",
        "list_test=os.listdir(path)\n",
        "image_test=[]\n",
        "for file in list_test:\n",
        "  list_=sorted(os.listdir(path+file))\n",
        "  image,_=load(path+file+\"/\"+list_[2])\n",
        "  for i in range(image.shape[2]):\n",
        "    image_test.append(image[:,:,i])\n",
        "    if i+1%100==0:\n",
        "      break"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ylTY9prN-JRl"
      },
      "source": [
        "images=np.reshape(images,(34100,240,240,1))\n",
        "labels=np.reshape(labels,(34100,240,240,1))"
      ],
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t_NS-VjK_Atj",
        "outputId": "7f4c2dc1-8353-414c-8e99-0210e9abcc2e",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "model.fit(images,labels,epochs=2,batch_size=32)\n",
        "model.save_weights(\"u_net_weight.h5\")"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/2\n",
            "   2/1066 [..............................] - ETA: 13:24 - loss: 0.0623 - acc: 0.9854WARNING:tensorflow:Callbacks method `on_train_batch_end` is slow compared to the batch time (batch time: 0.2736s vs `on_train_batch_end` time: 0.6196s). Check your callbacks.\n",
            "1066/1066 [==============================] - 952s 893ms/step - loss: 0.0672 - acc: 0.9843\n",
            "Epoch 2/2\n",
            "1066/1066 [==============================] - 952s 893ms/step - loss: 0.0587 - acc: 0.9790\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D1gQ7HnD_B2j",
        "outputId": "3f3bd342-533d-470c-cb15-1a73ae0596d9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        }
      },
      "source": [
        "image,_=load(\"/content/test/HGG_LGG/brats_2013_pat0134_1/VSD.Brain.XX.O.MR_T1c.54259.mha\")\n",
        "i=90\n",
        "plt.imshow(image[:,:,i])\n",
        "plt.imsave(\"image1.png\",image[:,:,i])"
      ],
      "execution_count": 127,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_QbBtYuT5IKV",
        "outputId": "71d73967-0f6c-4366-8141-a059b60f9b61",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        }
      },
      "source": [
        "image_pred=np.reshape(image[:,:,i],(1,240,240,1))\n",
        "image_pred=model.predict(image_pred)\n",
        "image_pred=np.reshape(image_pred,(240,240))\n",
        "plt.imshow(image_pred,\"gray\")\n",
        "plt.imsave(\"pred_image1.png\",image_pred,cmap=\"gray\")"
      ],
      "execution_count": 128,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ai-HKSzC5nlO"
      },
      "source": [
        "!cp \"/content/u_net_weight.h5\" -r \"/content/drive/My Drive\""
      ],
      "execution_count": 129,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TLZXMCML7QzJ"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}