mask_gen_matplotlib.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "from matplotlib import image\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>xray_fname</th>\n",
       "      <th>ctr</th>\n",
       "      <th>rc_x</th>\n",
       "      <th>rc_y</th>\n",
       "      <th>rh_x</th>\n",
       "      <th>rh_y</th>\n",
       "      <th>lh_x</th>\n",
       "      <th>lh_y</th>\n",
       "      <th>lc_x</th>\n",
       "      <th>lc_y</th>\n",
       "      <th>width</th>\n",
       "      <th>height</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>view1_frontal01.jpg</td>\n",
       "      <td>0.3114</td>\n",
       "      <td>320</td>\n",
       "      <td>1654</td>\n",
       "      <td>1214</td>\n",
       "      <td>1552</td>\n",
       "      <td>1860</td>\n",
       "      <td>1605</td>\n",
       "      <td>2395</td>\n",
       "      <td>1676</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>view1_frontal0100.jpg</td>\n",
       "      <td>0.5951</td>\n",
       "      <td>436</td>\n",
       "      <td>1735</td>\n",
       "      <td>1023</td>\n",
       "      <td>1566</td>\n",
       "      <td>2166</td>\n",
       "      <td>1708</td>\n",
       "      <td>2357</td>\n",
       "      <td>1725</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>view1_frontal0101.jpg</td>\n",
       "      <td>0.5317</td>\n",
       "      <td>329</td>\n",
       "      <td>1893</td>\n",
       "      <td>1010</td>\n",
       "      <td>1661</td>\n",
       "      <td>2113</td>\n",
       "      <td>1860</td>\n",
       "      <td>2404</td>\n",
       "      <td>1912</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>view1_frontal0103.jpg</td>\n",
       "      <td>0.5330</td>\n",
       "      <td>317</td>\n",
       "      <td>869</td>\n",
       "      <td>881</td>\n",
       "      <td>924</td>\n",
       "      <td>1635</td>\n",
       "      <td>1235</td>\n",
       "      <td>1731</td>\n",
       "      <td>881</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>view1_frontal0106.jpg</td>\n",
       "      <td>0.6726</td>\n",
       "      <td>838</td>\n",
       "      <td>1314</td>\n",
       "      <td>1306</td>\n",
       "      <td>1226</td>\n",
       "      <td>2448</td>\n",
       "      <td>1438</td>\n",
       "      <td>2536</td>\n",
       "      <td>1332</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>view1_frontal096.jpg</td>\n",
       "      <td>0.5991</td>\n",
       "      <td>564</td>\n",
       "      <td>1800</td>\n",
       "      <td>1431</td>\n",
       "      <td>1908</td>\n",
       "      <td>3000</td>\n",
       "      <td>2085</td>\n",
       "      <td>3183</td>\n",
       "      <td>1848</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>view1_frontal097.jpg</td>\n",
       "      <td>0.5931</td>\n",
       "      <td>688</td>\n",
       "      <td>1322</td>\n",
       "      <td>1276</td>\n",
       "      <td>1258</td>\n",
       "      <td>2232</td>\n",
       "      <td>1388</td>\n",
       "      <td>2300</td>\n",
       "      <td>1330</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>view1_frontal098.jpg</td>\n",
       "      <td>0.4263</td>\n",
       "      <td>126</td>\n",
       "      <td>1472</td>\n",
       "      <td>724</td>\n",
       "      <td>1428</td>\n",
       "      <td>1424</td>\n",
       "      <td>1554</td>\n",
       "      <td>1768</td>\n",
       "      <td>1500</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>200</th>\n",
       "      <td>view2_frontal054.jpg</td>\n",
       "      <td>0.4962</td>\n",
       "      <td>158</td>\n",
       "      <td>1282</td>\n",
       "      <td>612</td>\n",
       "      <td>1232</td>\n",
       "      <td>1396</td>\n",
       "      <td>1290</td>\n",
       "      <td>1738</td>\n",
       "      <td>1286</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>201</th>\n",
       "      <td>view2_frontal09.jpg</td>\n",
       "      <td>0.4541</td>\n",
       "      <td>128</td>\n",
       "      <td>1482</td>\n",
       "      <td>730</td>\n",
       "      <td>1482</td>\n",
       "      <td>1452</td>\n",
       "      <td>1542</td>\n",
       "      <td>1718</td>\n",
       "      <td>1476</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>202 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                xray_fname     ctr  rc_x  rc_y  rh_x  rh_y  lh_x  lh_y  lc_x  \\\n",
       "0      view1_frontal01.jpg  0.3114   320  1654  1214  1552  1860  1605  2395   \n",
       "1    view1_frontal0100.jpg  0.5951   436  1735  1023  1566  2166  1708  2357   \n",
       "2    view1_frontal0101.jpg  0.5317   329  1893  1010  1661  2113  1860  2404   \n",
       "3    view1_frontal0103.jpg  0.5330   317   869   881   924  1635  1235  1731   \n",
       "4    view1_frontal0106.jpg  0.6726   838  1314  1306  1226  2448  1438  2536   \n",
       "..                     ...     ...   ...   ...   ...   ...   ...   ...   ...   \n",
       "197   view1_frontal096.jpg  0.5991   564  1800  1431  1908  3000  2085  3183   \n",
       "198   view1_frontal097.jpg  0.5931   688  1322  1276  1258  2232  1388  2300   \n",
       "199   view1_frontal098.jpg  0.4263   126  1472   724  1428  1424  1554  1768   \n",
       "200   view2_frontal054.jpg  0.4962   158  1282   612  1232  1396  1290  1738   \n",
       "201    view2_frontal09.jpg  0.4541   128  1482   730  1482  1452  1542  1718   \n",
       "\n",
       "     lc_y  width  height  \n",
       "0    1676      0       0  \n",
       "1    1725      0       0  \n",
       "2    1912      0       0  \n",
       "3     881      0       0  \n",
       "4    1332      0       0  \n",
       "..    ...    ...     ...  \n",
       "197  1848      0       0  \n",
       "198  1330      0       0  \n",
       "199  1500      0       0  \n",
       "200  1286      0       0  \n",
       "201  1476      0       0  \n",
       "\n",
       "[202 rows x 12 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_df = pd.read_csv('CTR_Logs.txt', sep=' ', header=None, usecols=[0, 2, 4, 5, 7, 8, 10, 11, 13, 14])\n",
    "label_df.columns = ['xray_fname', 'ctr', 'rc_x', 'rc_y', 'rh_x', 'rh_y', 'lh_x', 'lh_y', 'lc_x', 'lc_y']\n",
    "label_cols = ['rc_x', 'rc_y', 'rh_x', 'rh_y', 'lh_x', 'lh_y', 'lc_x', 'lc_y']\n",
    "label_df[label_cols] = label_df[label_cols].round(0).astype(int)\n",
    "label_df['width'], label_df['height'] = 0, 0\n",
    "label_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2320, 2828)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "im = image.imread('./xray_images/view1_frontal01.jpg')\n",
    "print(im.shape)\n",
    "plt.imshow(im, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>xray_fname</th>\n",
       "      <th>ctr</th>\n",
       "      <th>rc_x</th>\n",
       "      <th>rc_y</th>\n",
       "      <th>rh_x</th>\n",
       "      <th>rh_y</th>\n",
       "      <th>lh_x</th>\n",
       "      <th>lh_y</th>\n",
       "      <th>lc_x</th>\n",
       "      <th>lc_y</th>\n",
       "      <th>width</th>\n",
       "      <th>height</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>view1_frontal01.jpg</td>\n",
       "      <td>0.3114</td>\n",
       "      <td>320</td>\n",
       "      <td>1654</td>\n",
       "      <td>1214</td>\n",
       "      <td>1552</td>\n",
       "      <td>1860</td>\n",
       "      <td>1605</td>\n",
       "      <td>2395</td>\n",
       "      <td>1676</td>\n",
       "      <td>2828</td>\n",
       "      <td>2320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>view1_frontal0100.jpg</td>\n",
       "      <td>0.5951</td>\n",
       "      <td>436</td>\n",
       "      <td>1735</td>\n",
       "      <td>1023</td>\n",
       "      <td>1566</td>\n",
       "      <td>2166</td>\n",
       "      <td>1708</td>\n",
       "      <td>2357</td>\n",
       "      <td>1725</td>\n",
       "      <td>2828</td>\n",
       "      <td>2320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>view1_frontal0101.jpg</td>\n",
       "      <td>0.5317</td>\n",
       "      <td>329</td>\n",
       "      <td>1893</td>\n",
       "      <td>1010</td>\n",
       "      <td>1661</td>\n",
       "      <td>2113</td>\n",
       "      <td>1860</td>\n",
       "      <td>2404</td>\n",
       "      <td>1912</td>\n",
       "      <td>2828</td>\n",
       "      <td>2320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>view1_frontal0103.jpg</td>\n",
       "      <td>0.5330</td>\n",
       "      <td>317</td>\n",
       "      <td>869</td>\n",
       "      <td>881</td>\n",
       "      <td>924</td>\n",
       "      <td>1635</td>\n",
       "      <td>1235</td>\n",
       "      <td>1731</td>\n",
       "      <td>881</td>\n",
       "      <td>2021</td>\n",
       "      <td>1631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>view1_frontal0106.jpg</td>\n",
       "      <td>0.6726</td>\n",
       "      <td>838</td>\n",
       "      <td>1314</td>\n",
       "      <td>1306</td>\n",
       "      <td>1226</td>\n",
       "      <td>2448</td>\n",
       "      <td>1438</td>\n",
       "      <td>2536</td>\n",
       "      <td>1332</td>\n",
       "      <td>2828</td>\n",
       "      <td>2320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>view1_frontal096.jpg</td>\n",
       "      <td>0.5991</td>\n",
       "      <td>564</td>\n",
       "      <td>1800</td>\n",
       "      <td>1431</td>\n",
       "      <td>1908</td>\n",
       "      <td>3000</td>\n",
       "      <td>2085</td>\n",
       "      <td>3183</td>\n",
       "      <td>1848</td>\n",
       "      <td>4240</td>\n",
       "      <td>3480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>view1_frontal097.jpg</td>\n",
       "      <td>0.5931</td>\n",
       "      <td>688</td>\n",
       "      <td>1322</td>\n",
       "      <td>1276</td>\n",
       "      <td>1258</td>\n",
       "      <td>2232</td>\n",
       "      <td>1388</td>\n",
       "      <td>2300</td>\n",
       "      <td>1330</td>\n",
       "      <td>2828</td>\n",
       "      <td>2320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>view1_frontal098.jpg</td>\n",
       "      <td>0.4263</td>\n",
       "      <td>126</td>\n",
       "      <td>1472</td>\n",
       "      <td>724</td>\n",
       "      <td>1428</td>\n",
       "      <td>1424</td>\n",
       "      <td>1554</td>\n",
       "      <td>1768</td>\n",
       "      <td>1500</td>\n",
       "      <td>1931</td>\n",
       "      <td>2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>200</th>\n",
       "      <td>view2_frontal054.jpg</td>\n",
       "      <td>0.4962</td>\n",
       "      <td>158</td>\n",
       "      <td>1282</td>\n",
       "      <td>612</td>\n",
       "      <td>1232</td>\n",
       "      <td>1396</td>\n",
       "      <td>1290</td>\n",
       "      <td>1738</td>\n",
       "      <td>1286</td>\n",
       "      <td>2022</td>\n",
       "      <td>1751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>201</th>\n",
       "      <td>view2_frontal09.jpg</td>\n",
       "      <td>0.4541</td>\n",
       "      <td>128</td>\n",
       "      <td>1482</td>\n",
       "      <td>730</td>\n",
       "      <td>1482</td>\n",
       "      <td>1452</td>\n",
       "      <td>1542</td>\n",
       "      <td>1718</td>\n",
       "      <td>1476</td>\n",
       "      <td>1721</td>\n",
       "      <td>2006</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>202 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                xray_fname     ctr  rc_x  rc_y  rh_x  rh_y  lh_x  lh_y  lc_x  \\\n",
       "0      view1_frontal01.jpg  0.3114   320  1654  1214  1552  1860  1605  2395   \n",
       "1    view1_frontal0100.jpg  0.5951   436  1735  1023  1566  2166  1708  2357   \n",
       "2    view1_frontal0101.jpg  0.5317   329  1893  1010  1661  2113  1860  2404   \n",
       "3    view1_frontal0103.jpg  0.5330   317   869   881   924  1635  1235  1731   \n",
       "4    view1_frontal0106.jpg  0.6726   838  1314  1306  1226  2448  1438  2536   \n",
       "..                     ...     ...   ...   ...   ...   ...   ...   ...   ...   \n",
       "197   view1_frontal096.jpg  0.5991   564  1800  1431  1908  3000  2085  3183   \n",
       "198   view1_frontal097.jpg  0.5931   688  1322  1276  1258  2232  1388  2300   \n",
       "199   view1_frontal098.jpg  0.4263   126  1472   724  1428  1424  1554  1768   \n",
       "200   view2_frontal054.jpg  0.4962   158  1282   612  1232  1396  1290  1738   \n",
       "201    view2_frontal09.jpg  0.4541   128  1482   730  1482  1452  1542  1718   \n",
       "\n",
       "     lc_y  width  height  \n",
       "0    1676   2828    2320  \n",
       "1    1725   2828    2320  \n",
       "2    1912   2828    2320  \n",
       "3     881   2021    1631  \n",
       "4    1332   2828    2320  \n",
       "..    ...    ...     ...  \n",
       "197  1848   4240    3480  \n",
       "198  1330   2828    2320  \n",
       "199  1500   1931    2011  \n",
       "200  1286   2022    1751  \n",
       "201  1476   1721    2006  \n",
       "\n",
       "[202 rows x 12 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the dimension of each xray file\n",
    "xray_images_path = './xray_images/'\n",
    "for i in range(label_df.shape[0]):\n",
    "    fname = label_df.xray_fname[i]\n",
    "    im = image.imread(os.path.join(xray_images_path, fname))\n",
    "    label_df.loc[i, 'height'], label_df.loc[i, 'width'] = im.shape\n",
    "\n",
    "label_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create masks\n",
    "out_path = './masks'\n",
    "\n",
    "if not os.path.exists(out_path):\n",
    "    os.makedirs(out_path)\n",
    "\n",
    "for i in range(label_df.shape[0]):\n",
    "    fname = label_df.xray_fname[i]\n",
    "    width = label_df.loc[i, 'width']\n",
    "    height = label_df.loc[i, 'height']\n",
    "    \n",
    "    mask_array = np.zeros((height, width), dtype=np.uint8)\n",
    "    \n",
    "    # horizontal x and vertical y with origin at top left\n",
    "    mask_array[label_df.rc_y[i], label_df.rc_x[i]] = 0.25 * 255\n",
    "    mask_array[label_df.rh_y[i], label_df.rh_x[i]] = 0.50 * 255\n",
    "    mask_array[label_df.lh_y[i], label_df.lh_x[i]] = 0.75 * 255\n",
    "    mask_array[label_df.lc_y[i], label_df.lc_x[i]] = 1.00 * 255\n",
    "\n",
    "    mask_path = os.path.join(out_path, fname[:-4]+'_mask.png')\n",
    "    plt.imsave(mask_path, mask_array, cmap=plt.cm.gray)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.75 0.75 0.75 1.  ]\n",
      "0.75\n",
      "(2320, 2828)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ii = 123\n",
    "fname = './masks/'+label_df.xray_fname[ii][:-4]+'_mask.png'\n",
    "# mask_example = image.imread(fname)\n",
    "mask_example = image.imread(fname).round(2)\n",
    "print(mask_example[label_df.lh_y[ii], label_df.lh_x[ii]])\n",
    "mask_example_squeeze = mask_example[:,:,:1].squeeze()\n",
    "print(mask_example_squeeze[label_df.lh_y[ii], label_df.lh_x[ii]])\n",
    "print(mask_example_squeeze.shape)\n",
    "plt.imshow(mask_example)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ii = 0\n",
    "fname = os.path.join('./xray_images', label_df.xray_fname[ii])\n",
    "im = image.imread(fname)\n",
    "im_copy = im.copy()\n",
    "# for i in range(label_df.rc_x[ii]-100, label_df.rc_x[ii]+100):\n",
    "#     for j in range(label_df.rc_y[ii]-100, label_df.rc_y[ii]+100):\n",
    "#         im_copy[i,j]=255\n",
    "# plt.imshow(im_copy, cmap='gray')\n",
    "plt.imshow(im, cmap='gray')\n",
    "plt.plot(label_df.rc_x[ii], label_df.rc_y[ii], 'bo')\n",
    "plt.plot(label_df.lc_x[ii], label_df.lc_y[ii], 'go')\n",
    "plt.plot(label_df.rh_x[ii], label_df.rh_y[ii], 'ro')\n",
    "plt.plot(label_df.lh_x[ii], label_df.lh_y[ii], 'mo')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot saliency map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import glob\n",
    "import torch\n",
    "import numpy as np\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision import transforms\n",
    "from torchvision.utils import save_image\n",
    "from cxr_mask_dataset import CXRMaskDataset\n",
    "from res_unet_model import ResnetUNet, UNet2\n",
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using cpu\n"
     ]
    }
   ],
   "source": [
    "OUTPUT_PATH = os.path.join('saliency_map')\n",
    "# DEVICE = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "DEVICE = torch.device('cpu')\n",
    "print(f'Using {DEVICE}')\n",
    "\n",
    "os.makedirs(OUTPUT_PATH, exist_ok=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_cxr_folder = os.path.join('data', 'processed', 'test', 'imgs')\n",
    "test_cxrs = glob.glob(os.path.join(test_cxr_folder, '*')) # A list of paths to all test cxr imgs\n",
    "\n",
    "cxr_transforms = transforms.Compose([\n",
    "            transforms.ToTensor(),\n",
    "            transforms.ConvertImageDtype(torch.float)\n",
    "        ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using cache found in C:\\Users\\Yehao/.cache\\torch\\hub\\milesial_Pytorch-UNet_master\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "UNet2(\n",
       "  (unet): UNet(\n",
       "    (inc): DoubleConv(\n",
       "      (double_conv): Sequential(\n",
       "        (0): Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "        (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (2): ReLU(inplace=True)\n",
       "        (3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "        (4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (5): ReLU(inplace=True)\n",
       "      )\n",
       "    )\n",
       "    (down1): Down(\n",
       "      (maxpool_conv): Sequential(\n",
       "        (0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
       "        (1): DoubleConv(\n",
       "          (double_conv): Sequential(\n",
       "            (0): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "            (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (2): ReLU(inplace=True)\n",
       "            (3): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "            (4): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (5): ReLU(inplace=True)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (down2): Down(\n",
       "      (maxpool_conv): Sequential(\n",
       "        (0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
       "        (1): DoubleConv(\n",
       "          (double_conv): Sequential(\n",
       "            (0): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "            (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (2): ReLU(inplace=True)\n",
       "            (3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "            (4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (5): ReLU(inplace=True)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (down3): Down(\n",
       "      (maxpool_conv): Sequential(\n",
       "        (0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
       "        (1): DoubleConv(\n",
       "          (double_conv): Sequential(\n",
       "            (0): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "            (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (2): ReLU(inplace=True)\n",
       "            (3): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "            (4): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (5): ReLU(inplace=True)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (down4): Down(\n",
       "      (maxpool_conv): Sequential(\n",
       "        (0): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
       "        (1): DoubleConv(\n",
       "          (double_conv): Sequential(\n",
       "            (0): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "            (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (2): ReLU(inplace=True)\n",
       "            (3): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "            (4): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "            (5): ReLU(inplace=True)\n",
       "          )\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (up1): Up(\n",
       "      (up): Upsample(scale_factor=2.0, mode=bilinear)\n",
       "      (conv): DoubleConv(\n",
       "        (double_conv): Sequential(\n",
       "          (0): Conv2d(1024, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "          (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU(inplace=True)\n",
       "          (3): Conv2d(512, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "          (4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (5): ReLU(inplace=True)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (up2): Up(\n",
       "      (up): Upsample(scale_factor=2.0, mode=bilinear)\n",
       "      (conv): DoubleConv(\n",
       "        (double_conv): Sequential(\n",
       "          (0): Conv2d(512, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU(inplace=True)\n",
       "          (3): Conv2d(256, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "          (4): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (5): ReLU(inplace=True)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (up3): Up(\n",
       "      (up): Upsample(scale_factor=2.0, mode=bilinear)\n",
       "      (conv): DoubleConv(\n",
       "        (double_conv): Sequential(\n",
       "          (0): Conv2d(256, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "          (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU(inplace=True)\n",
       "          (3): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "          (4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (5): ReLU(inplace=True)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (up4): Up(\n",
       "      (up): Upsample(scale_factor=2.0, mode=bilinear)\n",
       "      (conv): DoubleConv(\n",
       "        (double_conv): Sequential(\n",
       "          (0): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "          (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU(inplace=True)\n",
       "          (3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "          (4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (5): ReLU(inplace=True)\n",
       "        )\n",
       "      )\n",
       "    )\n",
       "    (outc): Conv2d(64, 3, kernel_size=(1, 1), stride=(1, 1))\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the model\n",
    "# resunet = ResnetUNet(3)\n",
    "# resunet.load_state_dict(torch.load(\"unet20.pt\"))\n",
    "resunet = UNet2()\n",
    "resunet.load_state_dict(torch.load('unet20.pt'))\n",
    "resunet.to(DEVICE)\n",
    "# print('model is on cuda: {resunet.is_cuda}')\n",
    "\n",
    "resunet.eval()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "bces = [0.564901, 0.483447, 0.451609, 0.414428, 0.380702, \\\n",
    "        0.350806, 0.322492, 0.296867, 0.273893, 0.252027, \\\n",
    "        0.231788, 0.213556, 0.197291, 0.182371, 0.168687, \\\n",
    "        0.155697, 0.143153, 0.132023, 0.122097, 0.113120]\n",
    "dices = [0.999991, 0.999990, 0.999989, 0.999988, 0.999988, \\\n",
    "        0.999987, 0.999986, 0.999985, 0.999984, 0.999982, \\\n",
    "        0.999981, 0.999979, 0.999978, 0.999976, 0.999975, \\\n",
    "        0.999973, 0.999971, 0.999968, 0.999966, 0.999963]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.782446,\n",
       " 0.7417185000000001,\n",
       " 0.725799,\n",
       " 0.7072080000000001,\n",
       " 0.690345,\n",
       " 0.6753965,\n",
       " 0.661239,\n",
       " 0.648426,\n",
       " 0.6369385,\n",
       " 0.6260045000000001,\n",
       " 0.6158845,\n",
       " 0.6067675,\n",
       " 0.5986345,\n",
       " 0.5911735,\n",
       " 0.5843309999999999,\n",
       " 0.577835,\n",
       " 0.571562,\n",
       " 0.5659955,\n",
       " 0.5610315,\n",
       " 0.5565415]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "losses = [0.5 * (bces[i] + dices[i]) for i in range(len(bces))]\n",
    "losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_epochs = range(1, len(losses)+1)\n",
    "\n",
    "plt.plot(n_epochs, bces, 'r', label='BCE Loss')\n",
    "plt.plot(n_epochs, dices, 'b', label='Dice Loss')\n",
    "plt.plot(n_epochs, losses, 'y', label = 'Weighted Loss')\n",
    "plt.title('BCE and Dice Loss')\n",
    "plt.legend()\n",
    "plt.savefig(f'unet_bce_dice_loss.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}