Prepare 3D CT Binary Classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [],
   "source": [
    "from _3D_CT_Preprocess import preprocessScan\n",
    "from _3D_CT_Preprocess import getScan\n",
    "from _3D_CT_Preprocess import normalizeScan\n",
    "from _3D_CT_Preprocess import resizeScan\n",
    "import numpy as np\n",
    "import SimpleITK as sitk\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy import ndimage\n",
    "import os\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "Cat = ['Positive','Negative']\n",
    "Count = [349,397]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BarContainer object of 2 artists>"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.bar(Cat,Count)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Direct Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "CovidScan=[]\n",
    "Label=[]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "CovidScan_Path_0 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-0/\"\n",
    "CovidScan_Path_1 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-1/\"\n",
    "CovidScan_Path_2 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-2/\"\n",
    "CovidScan_Path_3 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-3/\"\n",
    "CovidScan_Path_4 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-4/\"\n",
    "CovidScan_List_0 = os.listdir(CovidScan_Path_0)\n",
    "CovidScan_List_1 = os.listdir(CovidScan_Path_1)\n",
    "CovidScan_List_2 = os.listdir(CovidScan_Path_2)\n",
    "CovidScan_List_3 = os.listdir(CovidScan_Path_3)\n",
    "CovidScan_List_4 = os.listdir(CovidScan_Path_4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Negative CT\n",
    "for Scan_ in CovidScan_List_0:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    scan = preprocessScan(CovidScan_Path_0+Scan_,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(254, 254)"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Positive CT\n",
    "i = 0\n",
    "for Scan_ in CovidScan_List_1:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    scan = preprocessScan(CovidScan_Path_1+Scan_,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(1)\n",
    "    i=i+1\n",
    "    \n",
    "    if i == 103:\n",
    "        break;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "i=0\n",
    "for Scan_ in CovidScan_List_2:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    scan = preprocessScan(CovidScan_Path_2+Scan_,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(1)\n",
    "    i= i+1\n",
    "    \n",
    "    if i == 104:\n",
    "        break;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "for Scan_ in CovidScan_List_3:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    scan = preprocessScan(CovidScan_Path_3+Scan_,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(1)\n",
    "    i=i+1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(506, 506)"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "After Resize(128, 128, 64, 1)\n",
      "After Resize(128, 128, 64, 1)\n"
     ]
    }
   ],
   "source": [
    "for Scan_ in CovidScan_List_4:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    scan = preprocessScan(CovidScan_Path_4+Scan_,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(508, 508)"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(np.array(CovidScan), np.array(Label),test_size = 0.2,shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fc77283e070>"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.imshow(X_train[309][:,:,28],cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('X_train_3D_Direct_BinaryClass',X_train)\n",
    "np.save('y_train_3D_Direct_BinaryClass',y_train)\n",
    "np.save('X_test_3D_Direct_BinaryClass',X_test)\n",
    "np.save('y_test_3D_Direct_BinaryClass',y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((406, 128, 128, 64), (406,), (102, 128, 128, 64), (102,))"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape,y_train.shape,X_test.shape,y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Segmentation Lung to do classifiction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "CovidScan=[]\n",
    "Label=[]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "CovidScan_Path_0 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-0/\"\n",
    "CovidScan_Path_1 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-1/\"\n",
    "CovidScan_Path_2 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-2/\"\n",
    "CovidScan_Path_3 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-3/\"\n",
    "CovidScan_Path_4 = \"/Users/phoenixji/Desktop/LungData/MosMedData/CT-4/\"\n",
    "CovidScan_List_0 = os.listdir(CovidScan_Path_0)\n",
    "CovidScan_List_1 = os.listdir(CovidScan_Path_1)\n",
    "CovidScan_List_2 = os.listdir(CovidScan_Path_2)\n",
    "CovidScan_List_3 = os.listdir(CovidScan_Path_3)\n",
    "CovidScan_List_4 = os.listdir(CovidScan_Path_4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "255"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan_List_0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Negative CT\n",
    "for Scan_ in CovidScan_List_0:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    # Add the lung segmented CT to area\n",
    "    scan = getScan(CovidScan_Path_0+Scan_)\n",
    "    lungMask = np.array([get_segmented_lungs(slice.copy()) for slice in scan])\n",
    "    scan[~lungMask] = 0\n",
    "    scan = normalizeScan(scan)\n",
    "    print(\"Before Resize\"+str(scan.shape))\n",
    "    \n",
    "    if(scan.shape[0] != scan.shape[1]):\n",
    "        scan = np.swapaxes(scan,0,1)  #(64,128,128) => (128,128,64)\n",
    "        scan = np.swapaxes(scan,1,2)\n",
    "        \n",
    "    scan = resizeScan(scan,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(254, 254)"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Positive CT\n",
    "i = 0\n",
    "for Scan_ in CovidScan_List_1:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    # Add the lung segmented CT to area\n",
    "    scan = getScan(CovidScan_Path_1+Scan_)\n",
    "    lungMask = np.array([get_segmented_lungs(slice.copy()) for slice in scan])\n",
    "    scan[~lungMask] = 0\n",
    "    scan = normalizeScan(scan)\n",
    "    print(\"Before Resize\"+str(scan.shape))\n",
    "    \n",
    "    if(scan.shape[0] != scan.shape[1]):\n",
    "        scan = np.swapaxes(scan,0,1)  #(64,128,128) => (128,128,64)\n",
    "        scan = np.swapaxes(scan,1,2)\n",
    "        \n",
    "    scan = resizeScan(scan,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(1)\n",
    "    \n",
    "    i = i+1\n",
    "    if i==103:\n",
    "        break;\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(357, 357)"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "i=0\n",
    "for Scan_ in CovidScan_List_2:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    # Add the lung segmented CT to area\n",
    "    scan = getScan(CovidScan_Path_2+Scan_)\n",
    "    lungMask = np.array([get_segmented_lungs(slice.copy()) for slice in scan])\n",
    "    scan[~lungMask] = 0\n",
    "    scan = normalizeScan(scan)\n",
    "    print(\"Before Resize\"+str(scan.shape))\n",
    "    \n",
    "    if(scan.shape[0] != scan.shape[1]):\n",
    "        scan = np.swapaxes(scan,0,1)  #(64,128,128) => (128,128,64)\n",
    "        scan = np.swapaxes(scan,1,2)\n",
    "        \n",
    "    scan = resizeScan(scan,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(1)\n",
    "    \n",
    "    i = i+1\n",
    "    if i==103:\n",
    "        break;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(460, 460)"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "for Scan_ in CovidScan_List_3:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    # Add the lung segmented CT to area\n",
    "    scan = getScan(CovidScan_Path_3+Scan_)\n",
    "    lungMask = np.array([get_segmented_lungs(slice.copy()) for slice in scan])\n",
    "    scan[~lungMask] = 0\n",
    "    scan = normalizeScan(scan)\n",
    "    print(\"Before Resize\"+str(scan.shape))\n",
    "    \n",
    "    if(scan.shape[0] != scan.shape[1]):\n",
    "        scan = np.swapaxes(scan,0,1)  #(64,128,128) => (128,128,64)\n",
    "        scan = np.swapaxes(scan,1,2)\n",
    "        \n",
    "    scan = resizeScan(scan,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(1)\n",
    "    \n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(505, 505)"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Before Resize(42, 512, 512)\n",
      "After Resize(128, 128, 64, 1)\n",
      "Before Resize(36, 512, 512)\n",
      "After Resize(128, 128, 64, 1)\n"
     ]
    }
   ],
   "source": [
    "for Scan_ in CovidScan_List_4:\n",
    "    \n",
    "    if(Scan_ == \".DS_Store\"):\n",
    "        continue;\n",
    "    \n",
    "    # Add the lung segmented CT to area\n",
    "    scan = getScan(CovidScan_Path_4+Scan_)\n",
    "    lungMask = np.array([get_segmented_lungs(slice.copy()) for slice in scan])\n",
    "    scan[~lungMask] = 0\n",
    "    scan = normalizeScan(scan)\n",
    "    print(\"Before Resize\"+str(scan.shape))\n",
    "    \n",
    "    if(scan.shape[0] != scan.shape[1]):\n",
    "        scan = np.swapaxes(scan,0,1)  #(64,128,128) => (128,128,64)\n",
    "        scan = np.swapaxes(scan,1,2)\n",
    "        \n",
    "    scan = resizeScan(scan,(128,128,64))\n",
    "    scan = np.expand_dims(scan,axis=-1)\n",
    "    print(\"After Resize\"+str(scan.shape))\n",
    " \n",
    "    CovidScan.append(scan)\n",
    "    Label.append(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(507, 507)"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(CovidScan),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(128, 128, 64, 1)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CovidScan[0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fc7e8032b50>"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.imshow(CovidScan[-1][:,:,28],cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(np.array(CovidScan), np.array(Label), \n",
    "                                                    test_size = 0.2, random_state = 0,shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('X_train_3D_LungSeg_BinaryClass',X_train)\n",
    "np.save('y_train_3D_LungSeg_BinaryClass',y_train)\n",
    "np.save('X_test_3D_LungSeg_BinaryClass',X_test)\n",
    "np.save('y_test_3D_LungSeg_BinaryClass',y_test)"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}