masks_generation_dHCP.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import nibabel as nib\n",
    "from os.path import expanduser\n",
    "home = expanduser(\"~\")\n",
    "import os\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import glob\n",
    "import shutil\n",
    "#from PIL import Image\n",
    "#import uuid\n",
    "#import string, random\n",
    "import csv\n",
    "#import imageio\n",
    "#import tifffile as tif\n",
    "\n",
    "from os import listdir\n",
    "from os.path import splitext"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# put all the same type of image together in the same directory\n",
    "src_directory = home+'/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/data/'\n",
    "dst_directory = home+'/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/img_all/'\n",
    "dst_dir_mask = home+'/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/img_classification/ribbon_space-T2w_dseg/'\n",
    "dst_dir_imgs = home+'/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/img_classification/restore_T2w/'\n",
    "dst_dir_liq = home+'/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/img_classification/drawem9_space-T2w_dseg/'\n",
    "\n",
    "\n",
    "if not os.path.isdir(dst_directory):\n",
    "    os.mkdir(dst_directory)\n",
    "    \n",
    "if not os.path.isdir(dst_dir_mask):\n",
    "    os.mkdir(dst_dir_mask)\n",
    "    \n",
    "if not os.path.isdir(dst_dir_imgs):\n",
    "    os.mkdir(dst_dir_imgs)\n",
    "    \n",
    "if not os.path.isdir(dst_dir_liq):\n",
    "    os.mkdir(dst_dir_liq)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### obtain the MRI date of all the subjects"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "mri_directory = home+'/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/data/'\n",
    "\n",
    "#get directories for each subject\n",
    "mri_subjects = glob.glob(mri_directory + 'sub-*' +'/sub-*_sessions.tsv', recursive=True)\n",
    "# mri_sessions = glob.glob(mri_directory + 'sub-*' +'/ses-*', recursive=True)\n",
    "\n",
    "mri_list = []\n",
    "\n",
    "for i in range(len(mri_subjects)):\n",
    "    with open(mri_subjects[i], newline='') as tsvfile:\n",
    "        subject = mri_subjects[i].split('/')[10].split('-')[1]\n",
    "        # session = mri_sessions[i].split('/')[10].split('-')[1]\n",
    "        \n",
    "        reader_mri = csv.reader(tsvfile, delimiter='\\t')\n",
    "        data_mri = list(reader_mri)\n",
    "        data_mri.pop(0)\n",
    "        \n",
    "        for j in range(len(data_mri)):\n",
    "            #print(str(j)+'\\t'+str(data_mri))\n",
    "            vect = [subject, data_mri[j][0], round(float(data_mri[j][1]))]\n",
    "            mri_list.append(vect)\n",
    "    \n",
    "def take_third(elem):\n",
    "    return elem[2]\n",
    "    \n",
    "mri_date_s = sorted(mri_list, key=take_third)\n",
    "\n",
    "mri_date_20 = []\n",
    "mri_subj_20 = []\n",
    "count_mri = 0\n",
    "\n",
    "for i in range(len(mri_date_s)):\n",
    "    if count_mri < 600:\n",
    "        #if not any(mri_date_s[i][0] in s for s in mri_date_20):\n",
    "            # print(mri_date_s[i])\n",
    "        count_mri += 1\n",
    "        mri_date_20.append(mri_date_s[i])  # [\"id\", \"birth_date\"]\n",
    "        mri_subj_20.append(mri_date_s[i][0]+'_'+mri_date_s[i][1])\n",
    "            \n",
    "with open(dst_directory+'mri_date.csv', 'w', newline='') as csvfile:\n",
    "    writer = csv.writer(csvfile, delimiter='\\t')\n",
    "    writer.writerow(['Subject', 'Session', 'MRI date'])\n",
    "    for i in range(len(mri_date_20)):\n",
    "        writer.writerow(mri_date_20[i])\n",
    "        \n",
    "# print(sorted(mri_subj_20))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### obtain the birth date of all the subjects"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"# obtain birth date\n",
    "\n",
    "with open(src_directory+'participants.tsv', newline='') as tsvfile:\n",
    "    reader = csv.reader(tsvfile, delimiter='\\t')\n",
    "    data_birth = list(reader)\n",
    "    \n",
    "variables = data_birth[0]\n",
    "data_birth.pop(0)\n",
    "birth_date = []\n",
    "\n",
    "for row in data_birth:\n",
    "    vect = [row[0], row[2]] # [\"id\", \"birth_date\"]\n",
    "    birth_date.append(vect)\n",
    "    \n",
    "def take_second(elem):\n",
    "    return elem[1]\n",
    "    \n",
    "birth_date_s = sorted(birth_date, key=take_second)\n",
    "\n",
    "birth_date_20 = []\n",
    "count_birth = 0\n",
    "\n",
    "for i in range(len(birth_date_s)):\n",
    "    if float(birth_date_s[i][1]) < 30 and count_birth < 20:\n",
    "        count_birth += 1\n",
    "        birth_date_20.append(birth_date_s[i])  # [\"id\", \"birth_date\"]\n",
    "        \n",
    "# print(birth_date_20)\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### classify the files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "src_subjects = glob.glob(src_directory+'sub-*', recursive=True)\n",
    "num_cortex = 0\n",
    "    \n",
    "for subject in src_subjects:\n",
    "\n",
    "    sub = subject.split('/')[10].split('-')[1]        \n",
    "    src_sessions =  glob.glob(subject+'/ses-*', recursive=True)\n",
    "    \n",
    "    for session in src_sessions:\n",
    "    \n",
    "        ses = session.split('/')[11].split('-')[1]\n",
    "        name = sub+'_'+ses\n",
    "\n",
    "        if any(name in s for s in mri_subj_20): # birth_date_20\n",
    "\n",
    "            # from here the masks are obtained\n",
    "            src_cortex = glob.glob(session+'/*-ribbon_space-T2w_dseg.nii.gz')\n",
    "            if len(src_cortex) == 1:      \n",
    "                shutil.copy2(src_cortex[0], dst_dir_mask)\n",
    "\n",
    "                # from here the images are used if the corresponding image can be obtained\n",
    "                src_t2w = glob.glob(session+'/*-restore_T2w.nii.gz')[0]\n",
    "                shutil.copy2(src_t2w, dst_dir_imgs)\n",
    "\n",
    "                src_liq = glob.glob(session+'/*-drawem9_space-T2w_dseg.nii.gz')[0]\n",
    "                shutil.copy2(src_liq, dst_dir_liq)\n",
    "\n",
    "                num_cortex += 1\n",
    "                    \n",
    "img_t2w = sorted(glob.glob(dst_dir_imgs+'/*-restore_T2w.nii.gz'))\n",
    "img_cortex = sorted(glob.glob(dst_dir_mask+'/*-ribbon_space-T2w_dseg.nii.gz'))\n",
    "img_liq = sorted(glob.glob(dst_dir_liq+'/*-drawem9_space-T2w_dseg.nii.gz'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "img_t2w_ = sorted(glob.glob(dst_directory+'ribbon_space-T2w_dseg/imgs_3D/*'))\n",
    "img_cortex_ = sorted(glob.glob(dst_directory+'ribbon_space-T2w_dseg/mask_3D/*'))\n",
    "img_liq_ = sorted(glob.glob(dst_directory+'ribbon_space-T2w_dseg/total_mask_3D/*'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "img_t2w = sorted(glob.glob(dst_dir_imgs+'/*-restore_T2w.nii.gz'))\n",
    "img_cortex = sorted(glob.glob(dst_dir_mask+'/*-ribbon_space-T2w_dseg.nii.gz'))\n",
    "img_liq = sorted(glob.glob(dst_dir_liq+'/*-drawem9_space-T2w_dseg.nii.gz'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To generate the masks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mask for CC00060XX03-12501\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 1\n",
      "Subject: CC00060XX03 \t Session: 12501\n",
      "Mask for CC00062XX05-13801\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 2\n",
      "Subject: CC00062XX05 \t Session: 13801\n",
      "Mask for CC00063AN06-15102\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 3\n",
      "Subject: CC00063AN06 \t Session: 15102\n",
      "Mask for CC00065XX08-18600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 4\n",
      "Subject: CC00065XX08 \t Session: 18600\n",
      "Mask for CC00066XX09-19200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 5\n",
      "Subject: CC00066XX09 \t Session: 19200\n",
      "Mask for CC00067XX10-20200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 6\n",
      "Subject: CC00067XX10 \t Session: 20200\n",
      "Mask for CC00068XX11-20701\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 7\n",
      "Subject: CC00068XX11 \t Session: 20701\n",
      "Mask for CC00069XX12-26300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 8\n",
      "Subject: CC00069XX12 \t Session: 26300\n",
      "Mask for CC00071XX06-27000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 9\n",
      "Subject: CC00071XX06 \t Session: 27000\n",
      "Mask for CC00073XX08-27800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 10\n",
      "Subject: CC00073XX08 \t Session: 27800\n",
      "Mask for CC00074XX09-28000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 11\n",
      "Subject: CC00074XX09 \t Session: 28000\n",
      "Mask for CC00075XX10-28400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 12\n",
      "Subject: CC00075XX10 \t Session: 28400\n",
      "Mask for CC00078XX13-29900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 13\n",
      "Subject: CC00078XX13 \t Session: 29900\n",
      "Mask for CC00079XX14-30000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 14\n",
      "Subject: CC00079XX14 \t Session: 30000\n",
      "Mask for CC00080XX07-30300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 15\n",
      "Subject: CC00080XX07 \t Session: 30300\n",
      "Mask for CC00082XX09-30700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 16\n",
      "Subject: CC00082XX09 \t Session: 30700\n",
      "Mask for CC00083XX10-30900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 17\n",
      "Subject: CC00083XX10 \t Session: 30900\n",
      "Mask for CC00084XX11-31201\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 18\n",
      "Subject: CC00084XX11 \t Session: 31201\n",
      "Mask for CC00086XX13-31500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 19\n",
      "Subject: CC00086XX13 \t Session: 31500\n",
      "Mask for CC00087AN14-31800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 20\n",
      "Subject: CC00087AN14 \t Session: 31800\n",
      "Mask for CC00089XX16-32101\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 21\n",
      "Subject: CC00089XX16 \t Session: 32101\n",
      "Mask for CC00091XX10-32300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 22\n",
      "Subject: CC00091XX10 \t Session: 32300\n",
      "Mask for CC00094AN13-33500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 23\n",
      "Subject: CC00094AN13 \t Session: 33500\n",
      "Mask for CC00096XX15-33700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 24\n",
      "Subject: CC00096XX15 \t Session: 33700\n",
      "Mask for CC00099AN18-34200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 25\n",
      "Subject: CC00099AN18 \t Session: 34200\n",
      "Mask for CC00099BN18-34201\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 26\n",
      "Subject: CC00099BN18 \t Session: 34201\n",
      "Mask for CC00100XX01-35000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 27\n",
      "Subject: CC00100XX01 \t Session: 35000\n",
      "Mask for CC00101XX02-35001\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 28\n",
      "Subject: CC00101XX02 \t Session: 35001\n",
      "Mask for CC00102XX03-35200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 29\n",
      "Subject: CC00102XX03 \t Session: 35200\n",
      "Mask for CC00104XX05-35800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 30\n",
      "Subject: CC00104XX05 \t Session: 35800\n",
      "Mask for CC00106XX07-36100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 31\n",
      "Subject: CC00106XX07 \t Session: 36100\n",
      "Mask for CC00107XX08-36300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 32\n",
      "Subject: CC00107XX08 \t Session: 36300\n",
      "Mask for CC00108XX09-36800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 33\n",
      "Subject: CC00108XX09 \t Session: 36800\n",
      "Mask for CC00109XX10-36901\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 34\n",
      "Subject: CC00109XX10 \t Session: 36901\n",
      "Mask for CC00111XX04-37002\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 35\n",
      "Subject: CC00111XX04 \t Session: 37002\n",
      "Mask for CC00113XX06-37200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 36\n",
      "Subject: CC00113XX06 \t Session: 37200\n",
      "Mask for CC00114XX07-37100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 37\n",
      "Subject: CC00114XX07 \t Session: 37100\n",
      "Mask for CC00115XX08-37801\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 38\n",
      "Subject: CC00115XX08 \t Session: 37801\n",
      "Mask for CC00116XX09-38001\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 39\n",
      "Subject: CC00116XX09 \t Session: 38001\n",
      "Mask for CC00117XX10-38200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 40\n",
      "Subject: CC00117XX10 \t Session: 38200\n",
      "Mask for CC00119XX12-39400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 41\n",
      "Subject: CC00119XX12 \t Session: 39400\n",
      "Mask for CC00121XX06-41700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 42\n",
      "Subject: CC00121XX06 \t Session: 41700\n",
      "Mask for CC00122XX07-42000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 43\n",
      "Subject: CC00122XX07 \t Session: 42000\n",
      "Mask for CC00124XX09-42302\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 44\n",
      "Subject: CC00124XX09 \t Session: 42302\n",
      "Mask for CC00126XX11-43100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 45\n",
      "Subject: CC00126XX11 \t Session: 43100\n",
      "Mask for CC00127XX12-43200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 46\n",
      "Subject: CC00127XX12 \t Session: 43200\n",
      "Mask for CC00129BN14-43701\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 47\n",
      "Subject: CC00129BN14 \t Session: 43701\n",
      "Mask for CC00130XX07-44001\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 48\n",
      "Subject: CC00130XX07 \t Session: 44001\n",
      "Mask for CC00131XX08-44200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 49\n",
      "Subject: CC00131XX08 \t Session: 44200\n",
      "Mask for CC00132XX09-44400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 50\n",
      "Subject: CC00132XX09 \t Session: 44400\n",
      "Mask for CC00134XX11-44600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 51\n",
      "Subject: CC00134XX11 \t Session: 44600\n",
      "Mask for CC00135AN12-44900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 52\n",
      "Subject: CC00135AN12 \t Session: 44900\n",
      "Mask for CC00135AN12-54400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 53\n",
      "Subject: CC00135AN12 \t Session: 54400\n",
      "Mask for CC00136AN13-45100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 54\n",
      "Subject: CC00136AN13 \t Session: 45100\n",
      "Mask for CC00143AN12-47501\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 55\n",
      "Subject: CC00143AN12 \t Session: 47501\n",
      "Mask for CC00143BN12-47600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 56\n",
      "Subject: CC00143BN12 \t Session: 47600\n",
      "Mask for CC00144XX13-47701\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 57\n",
      "Subject: CC00144XX13 \t Session: 47701\n",
      "Mask for CC00145XX14-48100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 58\n",
      "Subject: CC00145XX14 \t Session: 48100\n",
      "Mask for CC00146XX15-48400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 59\n",
      "Subject: CC00146XX15 \t Session: 48400\n",
      "Mask for CC00149XX18-49000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 60\n",
      "Subject: CC00149XX18 \t Session: 49000\n",
      "Mask for CC00150AN02-54800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 61\n",
      "Subject: CC00150AN02 \t Session: 54800\n",
      "Mask for CC00150BN02-49100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 62\n",
      "Subject: CC00150BN02 \t Session: 49100\n",
      "Mask for CC00152AN04-49200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 63\n",
      "Subject: CC00152AN04 \t Session: 49200\n",
      "Mask for CC00153XX05-50300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 64\n",
      "Subject: CC00153XX05 \t Session: 50300\n",
      "Mask for CC00154XX06-50700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 65\n",
      "Subject: CC00154XX06 \t Session: 50700\n",
      "Mask for CC00157XX09-51900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 66\n",
      "Subject: CC00157XX09 \t Session: 51900\n",
      "Mask for CC00158XX10-52200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 67\n",
      "Subject: CC00158XX10 \t Session: 52200\n",
      "Mask for CC00160XX04-52700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 68\n",
      "Subject: CC00160XX04 \t Session: 52700\n",
      "Mask for CC00161XX05-53100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 69\n",
      "Subject: CC00161XX05 \t Session: 53100\n",
      "Mask for CC00168XX12-55700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 70\n",
      "Subject: CC00168XX12 \t Session: 55700\n",
      "Mask for CC00172AN08-56901\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 71\n",
      "Subject: CC00172AN08 \t Session: 56901\n",
      "Mask for CC00172BN08-57300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 72\n",
      "Subject: CC00172BN08 \t Session: 57300\n",
      "Mask for CC00174XX10-57200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 73\n",
      "Subject: CC00174XX10 \t Session: 57200\n",
      "Mask for CC00176XX12-57600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 74\n",
      "Subject: CC00176XX12 \t Session: 57600\n",
      "Mask for CC00177XX13-58500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 75\n",
      "Subject: CC00177XX13 \t Session: 58500\n",
      "Mask for CC00178XX14-58600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 76\n",
      "Subject: CC00178XX14 \t Session: 58600\n",
      "Mask for CC00179XX15-58800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 77\n",
      "Subject: CC00179XX15 \t Session: 58800\n",
      "Mask for CC00180XX08-59500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 78\n",
      "Subject: CC00180XX08 \t Session: 59500\n",
      "Mask for CC00181XX09-60000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 79\n",
      "Subject: CC00181XX09 \t Session: 60000\n",
      "Mask for CC00184XX12-60501\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 80\n",
      "Subject: CC00184XX12 \t Session: 60501\n",
      "Mask for CC00185XX13-60800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 81\n",
      "Subject: CC00185XX13 \t Session: 60800\n",
      "Mask for CC00186AN14-60900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 82\n",
      "Subject: CC00186AN14 \t Session: 60900\n",
      "Mask for CC00186BN14-61000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 83\n",
      "Subject: CC00186BN14 \t Session: 61000\n",
      "Mask for CC00189XX17-62301\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 84\n",
      "Subject: CC00189XX17 \t Session: 62301\n",
      "Mask for CC00192AN12-64301\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 85\n",
      "Subject: CC00192AN12 \t Session: 64301\n",
      "Mask for CC00193XX13-64400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 86\n",
      "Subject: CC00193XX13 \t Session: 64400\n",
      "Mask for CC00194XX14-65401\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 87\n",
      "Subject: CC00194XX14 \t Session: 65401\n",
      "Mask for CC00198XX18-66600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 88\n",
      "Subject: CC00198XX18 \t Session: 66600\n",
      "Mask for CC00199XX19-66700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 89\n",
      "Subject: CC00199XX19 \t Session: 66700\n",
      "Mask for CC00200XX02-67204\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1. 2. 3. 4.]\n",
      "Count: 90\n",
      "Subject: CC00200XX02 \t Session: 67204\n",
      "Mask for CC00201XX03-67600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 91\n",
      "Subject: CC00201XX03 \t Session: 67600\n",
      "Mask for CC00202XX04-67800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 92\n",
      "Subject: CC00202XX04 \t Session: 67800\n",
      "Mask for CC00203XX05-69500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 93\n",
      "Subject: CC00203XX05 \t Session: 69500\n",
      "Mask for CC00216AN10-73100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 94\n",
      "Subject: CC00216AN10 \t Session: 73100\n",
      "Mask for CC00217XX11-73700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 95\n",
      "Subject: CC00217XX11 \t Session: 73700\n",
      "Mask for CC00218AN12-74801\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 96\n",
      "Subject: CC00218AN12 \t Session: 74801\n",
      "Mask for CC00218BN12-74800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 97\n",
      "Subject: CC00218BN12 \t Session: 74800\n",
      "Mask for CC00219XX13-86600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 98\n",
      "Subject: CC00219XX13 \t Session: 86600\n",
      "Mask for CC00221XX07-75000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 99\n",
      "Subject: CC00221XX07 \t Session: 75000\n",
      "Mask for CC00223XX09-75900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 100\n",
      "Subject: CC00223XX09 \t Session: 75900\n",
      "Mask for CC00227XX13-92100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 101\n",
      "Subject: CC00227XX13 \t Session: 92100\n",
      "Mask for CC00231XX09-77701\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 102\n",
      "Subject: CC00231XX09 \t Session: 77701\n",
      "Mask for CC00236XX14-78700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 103\n",
      "Subject: CC00236XX14 \t Session: 78700\n",
      "Mask for CC00237XX15-79300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 104\n",
      "Subject: CC00237XX15 \t Session: 79300\n",
      "Mask for CC00238AN16-80300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 105\n",
      "Subject: CC00238AN16 \t Session: 80300\n",
      "Mask for CC00244XX14-81200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 106\n",
      "Subject: CC00244XX14 \t Session: 81200\n",
      "Mask for CC00245AN15-82304\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 107\n",
      "Subject: CC00245AN15 \t Session: 82304\n",
      "Mask for CC00245AN15-94300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 108\n",
      "Subject: CC00245AN15 \t Session: 94300\n",
      "Mask for CC00247XX17-82801\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 109\n",
      "Subject: CC00247XX17 \t Session: 82801\n",
      "Mask for CC00248XX18-92700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 110\n",
      "Subject: CC00248XX18 \t Session: 92700\n",
      "Mask for CC00254XX07-84300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 111\n",
      "Subject: CC00254XX07 \t Session: 84300\n",
      "Mask for CC00255XX08-84400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 112\n",
      "Subject: CC00255XX08 \t Session: 84400\n",
      "Mask for CC00257XX10-84700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 113\n",
      "Subject: CC00257XX10 \t Session: 84700\n",
      "Mask for CC00258XX11-84900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 114\n",
      "Subject: CC00258XX11 \t Session: 84900\n",
      "Mask for CC00260XX05-85300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 115\n",
      "Subject: CC00260XX05 \t Session: 85300\n",
      "Mask for CC00265XX10-86901\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 116\n",
      "Subject: CC00265XX10 \t Session: 86901\n",
      "Mask for CC00268XX13-87800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 117\n",
      "Subject: CC00268XX13 \t Session: 87800\n",
      "Mask for CC00269XX14-88300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 118\n",
      "Subject: CC00269XX14 \t Session: 88300\n",
      "Mask for CC00270XX07-88600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 119\n",
      "Subject: CC00270XX07 \t Session: 88600\n",
      "Mask for CC00271XX08-100400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 120\n",
      "Subject: CC00271XX08 \t Session: 100400\n",
      "Mask for CC00272XX09-117900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 121\n",
      "Subject: CC00272XX09 \t Session: 117900\n",
      "Mask for CC00281AN10-90400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 122\n",
      "Subject: CC00281AN10 \t Session: 90400\n",
      "Mask for CC00284AN13-111400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 123\n",
      "Subject: CC00284AN13 \t Session: 111400\n",
      "Mask for CC00284AN13-90701\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 124\n",
      "Subject: CC00284AN13 \t Session: 90701\n",
      "Mask for CC00286XX15-91700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 125\n",
      "Subject: CC00286XX15 \t Session: 91700\n",
      "Mask for CC00289XX18-119700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 126\n",
      "Subject: CC00289XX18 \t Session: 119700\n",
      "Mask for CC00292XX13-93800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 127\n",
      "Subject: CC00292XX13 \t Session: 93800\n",
      "Mask for CC00293AN14-97401\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 128\n",
      "Subject: CC00293AN14 \t Session: 97401\n",
      "Mask for CC00298XX19-94700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 129\n",
      "Subject: CC00298XX19 \t Session: 94700\n",
      "Mask for CC00300XX03-96000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 130\n",
      "Subject: CC00300XX03 \t Session: 96000\n",
      "Mask for CC00301XX04-113001\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 131\n",
      "Subject: CC00301XX04 \t Session: 113001\n",
      "Mask for CC00301XX04-96400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 132\n",
      "Subject: CC00301XX04 \t Session: 96400\n",
      "Mask for CC00302XX05-113500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 133\n",
      "Subject: CC00302XX05 \t Session: 113500\n",
      "Mask for CC00303XX06-96900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 134\n",
      "Subject: CC00303XX06 \t Session: 96900\n",
      "Mask for CC00304XX07-111600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 135\n",
      "Subject: CC00304XX07 \t Session: 111600\n",
      "Mask for CC00305XX08-115700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 136\n",
      "Subject: CC00305XX08 \t Session: 115700\n",
      "Mask for CC00305XX08-98101\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 137\n",
      "Subject: CC00305XX08 \t Session: 98101\n",
      "Mask for CC00306XX09-98700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 138\n",
      "Subject: CC00306XX09 \t Session: 98700\n",
      "Mask for CC00307XX10-98800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 139\n",
      "Subject: CC00307XX10 \t Session: 98800\n",
      "Mask for CC00308XX11-98900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 140\n",
      "Subject: CC00308XX11 \t Session: 98900\n",
      "Mask for CC00313XX08-100000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 141\n",
      "Subject: CC00313XX08 \t Session: 100000\n",
      "Mask for CC00314XX09-100101\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 142\n",
      "Subject: CC00314XX09 \t Session: 100101\n",
      "Mask for CC00316XX11-101300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 143\n",
      "Subject: CC00316XX11 \t Session: 101300\n",
      "Mask for CC00319XX14-117300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 144\n",
      "Subject: CC00319XX14 \t Session: 117300\n",
      "Mask for CC00320XX07-102300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 145\n",
      "Subject: CC00320XX07 \t Session: 102300\n",
      "Mask for CC00324XX11-111200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 146\n",
      "Subject: CC00324XX11 \t Session: 111200\n",
      "Mask for CC00326XX13-104100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 147\n",
      "Subject: CC00326XX13 \t Session: 104100\n",
      "Mask for CC00326XX13-118800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 148\n",
      "Subject: CC00326XX13 \t Session: 118800\n",
      "Mask for CC00329XX16-105400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 149\n",
      "Subject: CC00329XX16 \t Session: 105400\n",
      "Mask for CC00332XX11-105700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 150\n",
      "Subject: CC00332XX11 \t Session: 105700\n",
      "Mask for CC00334XX13-106100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 151\n",
      "Subject: CC00334XX13 \t Session: 106100\n",
      "Mask for CC00337XX16-107000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 152\n",
      "Subject: CC00337XX16 \t Session: 107000\n",
      "Mask for CC00338AN17-107700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 153\n",
      "Subject: CC00338AN17 \t Session: 107700\n",
      "Mask for CC00338BN17-107600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 154\n",
      "Subject: CC00338BN17 \t Session: 107600\n",
      "Mask for CC00339XX18-107200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 155\n",
      "Subject: CC00339XX18 \t Session: 107200\n",
      "Mask for CC00341XX12-108000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 156\n",
      "Subject: CC00341XX12 \t Session: 108000\n",
      "Mask for CC00342XX13-108300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 157\n",
      "Subject: CC00342XX13 \t Session: 108300\n",
      "Mask for CC00343XX14-108500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 158\n",
      "Subject: CC00343XX14 \t Session: 108500\n",
      "Mask for CC00344XX15-108600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 159\n",
      "Subject: CC00344XX15 \t Session: 108600\n",
      "Mask for CC00348XX19-110200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 160\n",
      "Subject: CC00348XX19 \t Session: 110200\n",
      "Mask for CC00349XX20-110300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 161\n",
      "Subject: CC00349XX20 \t Session: 110300\n",
      "Mask for CC00350XX04-109100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 162\n",
      "Subject: CC00350XX04 \t Session: 109100\n",
      "Mask for CC00351XX05-109500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 163\n",
      "Subject: CC00351XX05 \t Session: 109500\n",
      "Mask for CC00352XX06-110700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 164\n",
      "Subject: CC00352XX06 \t Session: 110700\n",
      "Mask for CC00364XX10-115200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 165\n",
      "Subject: CC00364XX10 \t Session: 115200\n",
      "Mask for CC00366XX12-116300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 166\n",
      "Subject: CC00366XX12 \t Session: 116300\n",
      "Mask for CC00367XX13-116000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 167\n",
      "Subject: CC00367XX13 \t Session: 116000\n",
      "Mask for CC00368XX14-116600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 168\n",
      "Subject: CC00368XX14 \t Session: 116600\n",
      "Mask for CC00376XX14-118400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 169\n",
      "Subject: CC00376XX14 \t Session: 118400\n",
      "Mask for CC00377XX15-119800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 170\n",
      "Subject: CC00377XX15 \t Session: 119800\n",
      "Mask for CC00382XX12-121700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 171\n",
      "Subject: CC00382XX12 \t Session: 121700\n",
      "Mask for CC00383XX13-121800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 172\n",
      "Subject: CC00383XX13 \t Session: 121800\n",
      "Mask for CC00384XX14-122200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 173\n",
      "Subject: CC00384XX14 \t Session: 122200\n",
      "Mask for CC00385XX15-118500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 174\n",
      "Subject: CC00385XX15 \t Session: 118500\n",
      "Mask for CC00385XX15-125700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 175\n",
      "Subject: CC00385XX15 \t Session: 125700\n",
      "Mask for CC00389XX19-119100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 176\n",
      "Subject: CC00389XX19 \t Session: 119100\n",
      "Mask for CC00389XX19-133800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 177\n",
      "Subject: CC00389XX19 \t Session: 133800\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mask for CC00398XX20-123100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 178\n",
      "Subject: CC00398XX20 \t Session: 123100\n",
      "Mask for CC00400XX04-123700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 179\n",
      "Subject: CC00400XX04 \t Session: 123700\n",
      "Mask for CC00406XX10-122900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 180\n",
      "Subject: CC00406XX10 \t Session: 122900\n",
      "Mask for CC00406XX10-129200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 181\n",
      "Subject: CC00406XX10 \t Session: 129200\n",
      "Mask for CC00407AN11-124000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 182\n",
      "Subject: CC00407AN11 \t Session: 124000\n",
      "Mask for CC00407AN11-137100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 183\n",
      "Subject: CC00407AN11 \t Session: 137100\n",
      "Mask for CC00408XX12-125500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 184\n",
      "Subject: CC00408XX12 \t Session: 125500\n",
      "Mask for CC00409XX13-125600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 185\n",
      "Subject: CC00409XX13 \t Session: 125600\n",
      "Mask for CC00414XX10-127300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 186\n",
      "Subject: CC00414XX10 \t Session: 127300\n",
      "Mask for CC00415XX11-127400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 187\n",
      "Subject: CC00415XX11 \t Session: 127400\n",
      "Mask for CC00416XX12-128900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 188\n",
      "Subject: CC00416XX12 \t Session: 128900\n",
      "Mask for CC00417XX13-129000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 189\n",
      "Subject: CC00417XX13 \t Session: 129000\n",
      "Mask for CC00418AN14-125200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 190\n",
      "Subject: CC00418AN14 \t Session: 125200\n",
      "Mask for CC00418AN14-130200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 191\n",
      "Subject: CC00418AN14 \t Session: 130200\n",
      "Mask for CC00418BN14-125300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 192\n",
      "Subject: CC00418BN14 \t Session: 125300\n",
      "Mask for CC00418BN14-130300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 193\n",
      "Subject: CC00418BN14 \t Session: 130300\n",
      "Mask for CC00421AN09-126000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 194\n",
      "Subject: CC00421AN09 \t Session: 126000\n",
      "Mask for CC00424XX12-129400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 195\n",
      "Subject: CC00424XX12 \t Session: 129400\n",
      "Mask for CC00428XX16-130400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 196\n",
      "Subject: CC00428XX16 \t Session: 130400\n",
      "Mask for CC00433XX13-132000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 197\n",
      "Subject: CC00433XX13 \t Session: 132000\n",
      "Mask for CC00434AN14-127700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 198\n",
      "Subject: CC00434AN14 \t Session: 127700\n",
      "Mask for CC00438XX18-156800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 199\n",
      "Subject: CC00438XX18 \t Session: 156800\n",
      "Mask for CC00439XX19-132100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 200\n",
      "Subject: CC00439XX19 \t Session: 132100\n",
      "Mask for CC00440XX12-132200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 201\n",
      "Subject: CC00440XX12 \t Session: 132200\n",
      "Mask for CC00441XX13-132902\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 202\n",
      "Subject: CC00441XX13 \t Session: 132902\n",
      "Mask for CC00442XX14-133300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 203\n",
      "Subject: CC00442XX14 \t Session: 133300\n",
      "Mask for CC00443XX15-133900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 204\n",
      "Subject: CC00443XX15 \t Session: 133900\n",
      "Mask for CC00445XX17-134800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 205\n",
      "Subject: CC00445XX17 \t Session: 134800\n",
      "Mask for CC00446XX18-135200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 206\n",
      "Subject: CC00446XX18 \t Session: 135200\n",
      "Mask for CC00447XX19-135600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 207\n",
      "Subject: CC00447XX19 \t Session: 135600\n",
      "Mask for CC00448XX20-135800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 208\n",
      "Subject: CC00448XX20 \t Session: 135800\n",
      "Mask for CC00451XX06-137000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 209\n",
      "Subject: CC00451XX06 \t Session: 137000\n",
      "Mask for CC00453XX08-136600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 210\n",
      "Subject: CC00453XX08 \t Session: 136600\n",
      "Mask for CC00454XX09-137600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 211\n",
      "Subject: CC00454XX09 \t Session: 137600\n",
      "Mask for CC00455XX10-137700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 212\n",
      "Subject: CC00455XX10 \t Session: 137700\n",
      "Mask for CC00457XX12-138601\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 213\n",
      "Subject: CC00457XX12 \t Session: 138601\n",
      "Mask for CC00461XX08-175100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 214\n",
      "Subject: CC00461XX08 \t Session: 175100\n",
      "Mask for CC00466BN13-138300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 215\n",
      "Subject: CC00466BN13 \t Session: 138300\n",
      "Mask for CC00468XX15-139100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 216\n",
      "Subject: CC00468XX15 \t Session: 139100\n",
      "Mask for CC00469XX16-139200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 217\n",
      "Subject: CC00469XX16 \t Session: 139200\n",
      "Mask for CC00470XX09-139600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 218\n",
      "Subject: CC00470XX09 \t Session: 139600\n",
      "Mask for CC00472XX11-140000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 219\n",
      "Subject: CC00472XX11 \t Session: 140000\n",
      "Mask for CC00478XX17-141601\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 220\n",
      "Subject: CC00478XX17 \t Session: 141601\n",
      "Mask for CC00479XX18-141608\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 221\n",
      "Subject: CC00479XX18 \t Session: 141608\n",
      "Mask for CC00480XX11-141609\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 222\n",
      "Subject: CC00480XX11 \t Session: 141609\n",
      "Mask for CC00481XX12-141800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 223\n",
      "Subject: CC00481XX12 \t Session: 141800\n",
      "Mask for CC00482XX13-142000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 224\n",
      "Subject: CC00482XX13 \t Session: 142000\n",
      "Mask for CC00483XX14-144200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 225\n",
      "Subject: CC00483XX14 \t Session: 144200\n",
      "Mask for CC00484XX15-142700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 226\n",
      "Subject: CC00484XX15 \t Session: 142700\n",
      "Mask for CC00486XX17-144300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 227\n",
      "Subject: CC00486XX17 \t Session: 144300\n",
      "Mask for CC00489XX20-138800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 228\n",
      "Subject: CC00489XX20 \t Session: 138800\n",
      "Mask for CC00489XX20-142900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 229\n",
      "Subject: CC00489XX20 \t Session: 142900\n",
      "Mask for CC00492AN15-140900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 230\n",
      "Subject: CC00492AN15 \t Session: 140900\n",
      "Mask for CC00497XX20-144500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 231\n",
      "Subject: CC00497XX20 \t Session: 144500\n",
      "Mask for CC00498XX21-144900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 232\n",
      "Subject: CC00498XX21 \t Session: 144900\n",
      "Mask for CC00500XX05-145900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 233\n",
      "Subject: CC00500XX05 \t Session: 145900\n",
      "Mask for CC00501XX06-146500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 234\n",
      "Subject: CC00501XX06 \t Session: 146500\n",
      "Mask for CC00502XX07-146700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 235\n",
      "Subject: CC00502XX07 \t Session: 146700\n",
      "Mask for CC00504XX09-146800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 236\n",
      "Subject: CC00504XX09 \t Session: 146800\n",
      "Mask for CC00505XX10-146900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 237\n",
      "Subject: CC00505XX10 \t Session: 146900\n",
      "Mask for CC00508XX13-148700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 238\n",
      "Subject: CC00508XX13 \t Session: 148700\n",
      "Mask for CC00512XX09-150700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 239\n",
      "Subject: CC00512XX09 \t Session: 150700\n",
      "Mask for CC00513XX10-150000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 240\n",
      "Subject: CC00513XX10 \t Session: 150000\n",
      "Mask for CC00518XX15-145700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 241\n",
      "Subject: CC00518XX15 \t Session: 145700\n",
      "Mask for CC00518XX15-166600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 242\n",
      "Subject: CC00518XX15 \t Session: 166600\n",
      "Mask for CC00520XX09-150201\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 243\n",
      "Subject: CC00520XX09 \t Session: 150201\n",
      "Mask for CC00525XX14-150600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 244\n",
      "Subject: CC00525XX14 \t Session: 150600\n",
      "Mask for CC00525XX14-165900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 245\n",
      "Subject: CC00525XX14 \t Session: 165900\n",
      "Mask for CC00526XX15-169900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 246\n",
      "Subject: CC00526XX15 \t Session: 169900\n",
      "Mask for CC00527XX16-184400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 247\n",
      "Subject: CC00527XX16 \t Session: 184400\n",
      "Mask for CC00528XX17-183200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 248\n",
      "Subject: CC00528XX17 \t Session: 183200\n",
      "Mask for CC00529AN18-151300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 249\n",
      "Subject: CC00529AN18 \t Session: 151300\n",
      "Mask for CC00529AN18-170000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 250\n",
      "Subject: CC00529AN18 \t Session: 170000\n",
      "Mask for CC00529BN18-151200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 251\n",
      "Subject: CC00529BN18 \t Session: 151200\n",
      "Mask for CC00529BN18-170100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 252\n",
      "Subject: CC00529BN18 \t Session: 170100\n",
      "Mask for CC00530XX11-153600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 253\n",
      "Subject: CC00530XX11 \t Session: 153600\n",
      "Mask for CC00532XX13-154200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 254\n",
      "Subject: CC00532XX13 \t Session: 154200\n",
      "Mask for CC00534XX15-155100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 255\n",
      "Subject: CC00534XX15 \t Session: 155100\n",
      "Mask for CC00536XX17-156900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 256\n",
      "Subject: CC00536XX17 \t Session: 156900\n",
      "Mask for CC00540XX13-164400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 257\n",
      "Subject: CC00540XX13 \t Session: 164400\n",
      "Mask for CC00542XX15-165800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 258\n",
      "Subject: CC00542XX15 \t Session: 165800\n",
      "Mask for CC00547XX20-157300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 259\n",
      "Subject: CC00547XX20 \t Session: 157300\n",
      "Mask for CC00548XX21-157400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 260\n",
      "Subject: CC00548XX21 \t Session: 157400\n",
      "Mask for CC00549XX22-157600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 261\n",
      "Subject: CC00549XX22 \t Session: 157600\n",
      "Mask for CC00550XX06-157800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 262\n",
      "Subject: CC00550XX06 \t Session: 157800\n",
      "Mask for CC00552XX08-159300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 263\n",
      "Subject: CC00552XX08 \t Session: 159300\n",
      "Mask for CC00553XX09-159000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1. 2. 3. 4.]\n",
      "Count: 264\n",
      "Subject: CC00553XX09 \t Session: 159000\n",
      "Mask for CC00556XX12-162902\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 265\n",
      "Subject: CC00556XX12 \t Session: 162902\n",
      "Mask for CC00558XX14-164100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 266\n",
      "Subject: CC00558XX14 \t Session: 164100\n",
      "Mask for CC00562XX10-191900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 267\n",
      "Subject: CC00562XX10 \t Session: 191900\n",
      "Mask for CC00563XX11-153900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 268\n",
      "Subject: CC00563XX11 \t Session: 153900\n",
      "Mask for CC00564XX12-154100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 269\n",
      "Subject: CC00564XX12 \t Session: 154100\n",
      "Mask for CC00566XX14-164500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 270\n",
      "Subject: CC00566XX14 \t Session: 164500\n",
      "Mask for CC00570XX10-158900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 271\n",
      "Subject: CC00570XX10 \t Session: 158900\n",
      "Mask for CC00571AN11-159101\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 272\n",
      "Subject: CC00571AN11 \t Session: 159101\n",
      "Mask for CC00572BN12-161400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 273\n",
      "Subject: CC00572BN12 \t Session: 161400\n",
      "Mask for CC00572CN12-161200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 274\n",
      "Subject: CC00572CN12 \t Session: 161200\n",
      "Mask for CC00576XX16-163200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 275\n",
      "Subject: CC00576XX16 \t Session: 163200\n",
      "Mask for CC00576XX16-178200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 276\n",
      "Subject: CC00576XX16 \t Session: 178200\n",
      "Mask for CC00577XX17-180400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 277\n",
      "Subject: CC00577XX17 \t Session: 180400\n",
      "Mask for CC00580XX12-173700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 278\n",
      "Subject: CC00580XX12 \t Session: 173700\n",
      "Mask for CC00581XX13-177101\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 279\n",
      "Subject: CC00581XX13 \t Session: 177101\n",
      "Mask for CC00582XX14-178300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 280\n",
      "Subject: CC00582XX14 \t Session: 178300\n",
      "Mask for CC00583XX15-178600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 281\n",
      "Subject: CC00583XX15 \t Session: 178600\n",
      "Mask for CC00588XX20-183600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 282\n",
      "Subject: CC00588XX20 \t Session: 183600\n",
      "Mask for CC00589XX21-184000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 283\n",
      "Subject: CC00589XX21 \t Session: 184000\n",
      "Mask for CC00590XX14-187000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 284\n",
      "Subject: CC00590XX14 \t Session: 187000\n",
      "Mask for CC00593XX17-189401\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 285\n",
      "Subject: CC00593XX17 \t Session: 189401\n",
      "Mask for CC00607XX13-179300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 286\n",
      "Subject: CC00607XX13 \t Session: 179300\n",
      "Mask for CC00613XX11-181601\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 287\n",
      "Subject: CC00613XX11 \t Session: 181601\n",
      "Mask for CC00616XX14-184700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 288\n",
      "Subject: CC00616XX14 \t Session: 184700\n",
      "Mask for CC00617XX15-176500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 289\n",
      "Subject: CC00617XX15 \t Session: 176500\n",
      "Mask for CC00617XX15-188400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 290\n",
      "Subject: CC00617XX15 \t Session: 188400\n",
      "Mask for CC00622XX12-185600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 291\n",
      "Subject: CC00622XX12 \t Session: 185600\n",
      "Mask for CC00628XX18-181800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 292\n",
      "Subject: CC00628XX18 \t Session: 181800\n",
      "Mask for CC00628XX18-193500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 293\n",
      "Subject: CC00628XX18 \t Session: 193500\n",
      "Mask for CC00629XX19-198100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 294\n",
      "Subject: CC00629XX19 \t Session: 198100\n",
      "Mask for CC00632XX14-183300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 295\n",
      "Subject: CC00632XX14 \t Session: 183300\n",
      "Mask for CC00639XX21-216800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 296\n",
      "Subject: CC00639XX21 \t Session: 216800\n",
      "Mask for CC00647XX21-190801\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 297\n",
      "Subject: CC00647XX21 \t Session: 190801\n",
      "Mask for CC00650XX07-218007\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 298\n",
      "Subject: CC00650XX07 \t Session: 218007\n",
      "Mask for CC00652XX09-191600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 299\n",
      "Subject: CC00652XX09 \t Session: 191600\n",
      "Mask for CC00653XX10-191801\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 300\n",
      "Subject: CC00653XX10 \t Session: 191801\n",
      "Mask for CC00654XX11-192000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 301\n",
      "Subject: CC00654XX11 \t Session: 192000\n",
      "Mask for CC00656XX13-217601\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 302\n",
      "Subject: CC00656XX13 \t Session: 217601\n",
      "Mask for CC00657XX14-193700\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 303\n",
      "Subject: CC00657XX14 \t Session: 193700\n",
      "Mask for CC00657XX14-212200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 304\n",
      "Subject: CC00657XX14 \t Session: 212200\n",
      "Mask for CC00661XX10-209600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 305\n",
      "Subject: CC00661XX10 \t Session: 209600\n",
      "Mask for CC00667XX16-196800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 306\n",
      "Subject: CC00667XX16 \t Session: 196800\n",
      "Mask for CC00669XX18-214300\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 307\n",
      "Subject: CC00669XX18 \t Session: 214300\n",
      "Mask for CC00670XX11-197200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 308\n",
      "Subject: CC00670XX11 \t Session: 197200\n",
      "Mask for CC00672BN13-200000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 309\n",
      "Subject: CC00672BN13 \t Session: 200000\n",
      "Mask for CC00672BN13-214800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 310\n",
      "Subject: CC00672BN13 \t Session: 214800\n",
      "Mask for CC00675XX16-219100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 311\n",
      "Subject: CC00675XX16 \t Session: 219100\n",
      "Mask for CC00685XX18-226100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 312\n",
      "Subject: CC00685XX18 \t Session: 226100\n",
      "Mask for CC00686XX19-198800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 313\n",
      "Subject: CC00686XX19 \t Session: 198800\n",
      "Mask for CC00686XX19-209100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 314\n",
      "Subject: CC00686XX19 \t Session: 209100\n",
      "Mask for CC00687XX20-199000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 315\n",
      "Subject: CC00687XX20 \t Session: 199000\n",
      "Mask for CC00688XX21-199500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 316\n",
      "Subject: CC00688XX21 \t Session: 199500\n",
      "Mask for CC00688XX21-205500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 317\n",
      "Subject: CC00688XX21 \t Session: 205500\n",
      "Mask for CC00693XX18-201000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 318\n",
      "Subject: CC00693XX18 \t Session: 201000\n",
      "Mask for CC00695XX20-202600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 319\n",
      "Subject: CC00695XX20 \t Session: 202600\n",
      "Mask for CC00702AN09-204600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 320\n",
      "Subject: CC00702AN09 \t Session: 204600\n",
      "Mask for CC00703XX10-206101\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 321\n",
      "Subject: CC00703XX10 \t Session: 206101\n",
      "Mask for CC00712XX11-232701\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 322\n",
      "Subject: CC00712XX11 \t Session: 232701\n",
      "Mask for CC00719XX18-210600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 323\n",
      "Subject: CC00719XX18 \t Session: 210600\n",
      "Mask for CC00720XX11-211101\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 324\n",
      "Subject: CC00720XX11 \t Session: 211101\n",
      "Mask for CC00723XX14-211900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 325\n",
      "Subject: CC00723XX14 \t Session: 211900\n",
      "Mask for CC00723XX14-224100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 326\n",
      "Subject: CC00723XX14 \t Session: 224100\n",
      "Mask for CC00731XX14-214500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 327\n",
      "Subject: CC00731XX14 \t Session: 214500\n",
      "Mask for CC00734XX17-216900\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 328\n",
      "Subject: CC00734XX17 \t Session: 216900\n",
      "Mask for CC00735XX18-222201\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 329\n",
      "Subject: CC00735XX18 \t Session: 222201\n",
      "Mask for CC00740XX15-238400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 330\n",
      "Subject: CC00740XX15 \t Session: 238400\n",
      "Mask for CC00747XX22-238600\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 331\n",
      "Subject: CC00747XX22 \t Session: 238600\n",
      "Mask for CC00754AN12-223800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 332\n",
      "Subject: CC00754AN12 \t Session: 223800\n",
      "Mask for CC00754BN12-223701\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 333\n",
      "Subject: CC00754BN12 \t Session: 223701\n",
      "Mask for CC00757XX15-12010\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 334\n",
      "Subject: CC00757XX15 \t Session: 12010\n",
      "Mask for CC00760XX10-229500\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 335\n",
      "Subject: CC00760XX10 \t Session: 229500\n",
      "Mask for CC00764AN14-231800\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 336\n",
      "Subject: CC00764AN14 \t Session: 231800\n",
      "Mask for CC00764BN14-231702\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 337\n",
      "Subject: CC00764BN14 \t Session: 231702\n",
      "Mask for CC00765XX15-8210\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 338\n",
      "Subject: CC00765XX15 \t Session: 8210\n",
      "Mask for CC00766XX16-4200\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 339\n",
      "Subject: CC00766XX16 \t Session: 4200\n",
      "Mask for CC00770XX12-1100\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 340\n",
      "Subject: CC00770XX12 \t Session: 1100\n",
      "Mask for CC00770XX12-238000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 341\n",
      "Subject: CC00770XX12 \t Session: 238000\n",
      "Mask for CC00771XX13-17710\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 342\n",
      "Subject: CC00771XX13 \t Session: 17710\n",
      "Mask for CC00777XX19-239102\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 343\n",
      "Subject: CC00777XX19 \t Session: 239102\n",
      "Mask for CC00788XX22-243000\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 344\n",
      "Subject: CC00788XX22 \t Session: 243000\n",
      "Mask for CC00797XX23-12110\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 345\n",
      "Subject: CC00797XX23 \t Session: 12110\n",
      "Mask for CC00797XX23-6210\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 346\n",
      "Subject: CC00797XX23 \t Session: 6210\n",
      "Mask for CC00798XX24-245400\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 347\n",
      "Subject: CC00798XX24 \t Session: 245400\n",
      "Mask for CC00799XX25-23810\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 348\n",
      "Subject: CC00799XX25 \t Session: 23810\n",
      "Mask for CC00801XX09-29510\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 349\n",
      "Subject: CC00801XX09 \t Session: 29510\n",
      "Mask for CC00804XX12-9310\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 350\n",
      "Subject: CC00804XX12 \t Session: 9310\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mask for CC00805XX13-24210\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 351\n",
      "Subject: CC00805XX13 \t Session: 24210\n",
      "Mask for CC00815XX15-4120\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 352\n",
      "Subject: CC00815XX15 \t Session: 4120\n",
      "Mask for CC00818XX18-4020\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 353\n",
      "Subject: CC00818XX18 \t Session: 4020\n",
      "Mask for CC00822XX14-15710\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 354\n",
      "Subject: CC00822XX14 \t Session: 15710\n",
      "Mask for CC00823XX15-15810\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 355\n",
      "Subject: CC00823XX15 \t Session: 15810\n",
      "Mask for CC00824XX16-38310\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 356\n",
      "Subject: CC00824XX16 \t Session: 38310\n",
      "Mask for CC00830XX14-30710\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 357\n",
      "Subject: CC00830XX14 \t Session: 30710\n",
      "Mask for CC00833XX17-20510\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 358\n",
      "Subject: CC00833XX17 \t Session: 20510\n",
      "Mask for CC00838XX22-21910\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 359\n",
      "Subject: CC00838XX22 \t Session: 21910\n",
      "Mask for CC00838XX22-30610\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 360\n",
      "Subject: CC00838XX22 \t Session: 30610\n",
      "Mask for CC00840XX16-24910\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 361\n",
      "Subject: CC00840XX16 \t Session: 24910\n",
      "Mask for CC00843XX19-4330\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 362\n",
      "Subject: CC00843XX19 \t Session: 4330\n",
      "Mask for CC00845AN21-26510\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 363\n",
      "Subject: CC00845AN21 \t Session: 26510\n",
      "Mask for CC00845AN21-32010\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 364\n",
      "Subject: CC00845AN21 \t Session: 32010\n",
      "Mask for CC00845BN21-26410\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 365\n",
      "Subject: CC00845BN21 \t Session: 26410\n",
      "Mask for CC00845BN21-32110\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 366\n",
      "Subject: CC00845BN21 \t Session: 32110\n",
      "Mask for CC00850XX09-4930\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 367\n",
      "Subject: CC00850XX09 \t Session: 4930\n",
      "Mask for CC00852XX11-28210\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 368\n",
      "Subject: CC00852XX11 \t Session: 28210\n",
      "Mask for CC00858XX17-32210\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 369\n",
      "Subject: CC00858XX17 \t Session: 32210\n",
      "Mask for CC00860XX11-7030\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 370\n",
      "Subject: CC00860XX11 \t Session: 7030\n",
      "Mask for CC00863XX14-34810\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 371\n",
      "Subject: CC00863XX14 \t Session: 34810\n",
      "Mask for CC00871XX14-38810\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 372\n",
      "Subject: CC00871XX14 \t Session: 38810\n",
      "Mask for CC00907XX16-4230\n",
      "[0. 1. 2. 3. 4.]\n",
      "Count: 373\n",
      "Subject: CC00907XX16 \t Session: 4230\n"
     ]
    }
   ],
   "source": [
    "count = 0\n",
    "\n",
    "dirr = img_cortex\n",
    "\n",
    "for j in range(len(dirr)):\n",
    "    sub = dirr[j].split('/')[11].split('_')[0].split('-')[1]\n",
    "    ses = dirr[j].split('/')[11].split('_')[1].split('-')[1]\n",
    "\n",
    "    sub_liq = img_liq[j].split('/')[11].split('_')[0].split('-')[1]\n",
    "    ses_liq = img_liq[j].split('/')[11].split('_')[1].split('-')[1]\n",
    "\n",
    "    img_ = nib.load(dirr[j])\n",
    "    data_m = img_.get_fdata()\n",
    "    data_liq = nib.load(img_liq[j]).get_fdata()\n",
    "\n",
    "    mask_cp = np.random.rand(*data_m.shape)\n",
    "    mask_t = np.random.rand(*data_m.shape)\n",
    "\n",
    "    \"\"\"# mask for cortical plate\n",
    "    print('CP mask for '+str(sub)+'-'+str(ses))\n",
    "    for i in range(len(data_m[1,1,:])):\n",
    "        # values of the cortical plate\n",
    "        mask_cp[:,:,i] = (data_m[:,:,i] == 42)\n",
    "        mask_cp[:,:,i] += (data_m[:,:,i] == 3)\"\"\"\n",
    "    \n",
    "    # 2, 3, 1+4+5, 7+9, 6+8\n",
    "    \n",
    "    print('Mask for '+str(sub)+'-'+str(ses))\n",
    "    for i in range(len(data_m[1,1,:])):\n",
    "        # values of the cortical plate\n",
    "        mask_t[:,:,i] = np.where(data_liq[:,:,i] == 1, 1, 0)\n",
    "        mask_t[:,:,i] += np.where(data_liq[:,:,i] == 5, 1, 0)\n",
    "        mask_t[:,:,i] += np.where(data_liq[:,:,i] == 2, 2, 0)\n",
    "        mask_t[:,:,i] += np.where(data_liq[:,:,i] == 3, 3, 0)\n",
    "        mask_t[:,:,i] += np.where(data_liq[:,:,i] == 7, 4, 0)\n",
    "        \n",
    "    print(np.unique(mask_t))\n",
    "\n",
    "    \"\"\"    if sub == sub_liq:\n",
    "        if ses == ses_liq:\n",
    "            print('Total mask for '+str(sub)+'-'+str(ses))\n",
    "            #mask for the whole brain\n",
    "            for i in range(len(data_m[1,1,:])):\n",
    "                #mask_t[:,:,i] = (data_m[:,:,i] == 42)\n",
    "                #mask_t[:,:,i] += (data_m[:,:,i] == 3)\n",
    "                #mask_t[:,:,i] += (data_m[:,:,i] == 2)\n",
    "                #mask_t[:,:,i] += (data_m[:,:,i] == 41)\n",
    "                mask_t[:,:,i] += (data_liq[:,:,i] == 1)\n",
    "                mask_t[:,:,i] += (data_liq[:,:,i] == 2)\n",
    "                mask_t[:,:,i] += (data_liq[:,:,i] == 3)\n",
    "                mask_t[:,:,i] += (data_liq[:,:,i] == 5)\n",
    "                mask_t[:,:,i] += (data_liq[:,:,i] == 6)\n",
    "                mask_t[:,:,i] += (data_liq[:,:,i] == 7)\n",
    "                mask_t[:,:,i] += (data_liq[:,:,i] == 8)\n",
    "                mask_t[:,:,i] += (data_liq[:,:,i] == 9)\"\"\"\n",
    "\n",
    "    #threshold = 1\n",
    "    #mask_t = np.where(mask_t>threshold, 1, 0)\n",
    "\n",
    "    for k in range(len(img_t2w)):\n",
    "        sub_ = img_t2w[k].split('/')[11].split('_')[0].split('-')[1]\n",
    "\n",
    "        if sub == sub_:\n",
    "\n",
    "            ses_ = img_t2w[k].split('/')[11].split('_')[1].split('-')[1]\n",
    "\n",
    "            if ses == ses_:\n",
    "\n",
    "                mask_ = mask_t.astype(np.float32)\n",
    "                mask_cp = nib.Nifti1Image(mask_, img_.affine, img_.header)\n",
    "\n",
    "                #mask_tt = mask_t.astype(np.float32)\n",
    "                #mask_nt = nib.Nifti1Image(mask_tt, img_.affine, img_.header)\n",
    "\n",
    "                #data_ = data_m.astype(np.float32)\n",
    "                #data_n = nib.Nifti1Image(data_, img_.affine, img_.header)\n",
    "\n",
    "                #nib.save(data_n, dst_directory+'ribbon_space-T2w_dseg/imgs_3D/'+str(sub)+'_'+str(ses)+'.nii.gz')\n",
    "                #nib.save(mask_cp, dst_directory+'ribbon_space-T2w_dseg/mask_3D/'+str(sub)+'_'+str(ses)+'.nii.gz')\n",
    "                nib.save(mask_cp, dst_directory+'ribbon_space-T2w_dseg/mask_3D_multi_v4/'+str(sub)+'_'+str(ses)+'.nii.gz')\n",
    "                #nib.save(mask_nt, dst_directory+'ribbon_space-T2w_dseg/total_mask_3D/'+str(sub)+'_'+str(ses)+'.nii.gz')\n",
    "\n",
    "                # copy the corresponding image of the mask with the same name\n",
    "                #shutil.copy2(img_t2w[k], dst_directory+'restore_T2w/'+str(sub)+'_'+str(ses)+'.nii.gz')\n",
    "\n",
    "                count += 1\n",
    "\n",
    "                print('Count: '+str(count))\n",
    "                print('Subject: '+str(sub)+' \\t Session: '+str(ses))\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To take the N youngest subjects"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0: \tCC00530XX11_153600:\t29\n",
      "1: \tCC00735XX18_222201:\t29\n",
      "2: \tCC00389XX19_119100:\t30\n",
      "3: \tCC00657XX14_193700:\t30\n",
      "4: \tCC00618XX16_177201:\t30\n",
      "4: \tCC00672AN13_197601:\t30\n",
      "4: \tCC00723XX14_211900:\t31\n",
      "5: \tCC00672BN13_200000:\t31\n",
      "6: \tCC00284AN13_90701:\t32\n",
      "7: \tCC00284BN13_90801:\t32\n",
      "7: \tCC00621XX11_177900:\t32\n",
      "7: \tCC00136AN13_45100:\t33\n",
      "8: \tCC00518XX15_145700:\t33\n",
      "9: \tCC00760XX10_229500:\t33\n",
      "10: \tCC00301XX04_96400:\t33\n",
      "11: \tCC00529AN18_151300:\t33\n",
      "12: \tCC00838XX22_21910:\t33\n",
      "13: \tCC00686XX19_198800:\t33\n",
      "14: \tCC00823XX15_15810:\t33\n",
      "15: \tCC00529BN18_151200:\t33\n",
      "16: \tCC00305XX08_98101:\t33\n",
      "17: \tCC00829XX21_17610:\t33\n",
      "17: \tCC00578AN18_164900:\t33\n",
      "17: \tCC00525XX14_150600:\t34\n",
      "18: \tCC00154XX06_50700:\t34\n",
      "19: \tCC00293AN14_97401:\t34\n",
      "20: \tCC00628XX18_181800:\t34\n",
      "21: \tCC00124XX09_42302:\t34\n",
      "22: \tCC00703XX10_206101:\t34\n",
      "23: \tCC00617XX15_176500:\t34\n",
      "24: \tCC00788XX22_243000:\t34\n",
      "25: \tCC00293BN14_97500:\t34\n",
      "25: \tCC00423XX11_127200:\t34\n",
      "25: \tCC00135AN12_44900:\t35\n",
      "26: \tCC00571AN11_159101:\t35\n",
      "27: \tCC00245AN15_82304:\t35\n",
      "28: \tCC00563XX11_153900:\t35\n",
      "29: \tCC00632XX14_183300:\t35\n",
      "30: \tCC00804XX12_900:\t35\n",
      "30: \tCC00218AN12_74801:\t35\n",
      "31: \tCC00351XX05_109500:\t35\n",
      "32: \tCC00770XX12_238000:\t35\n",
      "33: \tCC00177XX13_58500:\t35\n",
      "34: \tCC00152AN04_49200:\t35\n",
      "35: \tCC00218BN12_74800:\t35\n",
      "36: \tCC00907XX16_4230:\t35\n",
      "37: \tCC00792XX18_244200:\t35\n",
      "37: \tCC00245BN15_82600:\t35\n",
      "37: \tCC00600XX06_175800:\t35\n",
      "37: \tCC00689XX22_199800:\t35\n",
      "37: \tCC00135BN12_44704:\t35\n",
      "37: \tCC00191XX11_65200:\t35\n",
      "37: \tCC00569XX17_158300:\t35\n",
      "37: \tCC00889BN24_2320:\t35\n",
      "37: \tCC00147XX16_48500:\t35\n",
      "37: \tCC00407AN11_124000:\t36\n",
      "38: \tCC00570XX10_158900:\t36\n",
      "39: \tCC00406XX10_122900:\t36\n"
     ]
    }
   ],
   "source": [
    "mri_list_ = []\n",
    "num_subj = 40\n",
    "train_set = '/Users/meritxell/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/training_set/'\n",
    "\n",
    "with open('/Users/meritxell/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/img_all/mri_date.csv', newline='') as csvfile:\n",
    "    reader = csv.reader(csvfile, delimiter='\\t')\n",
    "    for row in reader:\n",
    "        mri_list_.append(row)\n",
    "        \n",
    "mri_list_.pop(0)\n",
    "count = 0\n",
    "\n",
    "for i in range(60):\n",
    "    if count < num_subj:\n",
    "        idx = mri_list_[i][0]+'_'+mri_list_[i][1]\n",
    "        print(str(count)+': \\t'+idx+':\\t'+mri_list_[i][2])\n",
    "\n",
    "        img_t2w_ = dst_directory+'restore_T2w/'+idx+'.nii.gz'\n",
    "        #img_cortex_ = dst_directory+'ribbon_space-T2w_dseg/mask_3D/'+idx+'.nii.gz'\n",
    "        img_cortex_v2_ = dst_directory+'ribbon_space-T2w_dseg/mask_3D_multi_v4/'+idx+'.nii.gz'\n",
    "        #img_liq_ = dst_directory+'ribbon_space-T2w_dseg/total_mask_3D/'+idx+'.nii.gz'\n",
    "        file = glob.glob(img_t2w_)\n",
    "\n",
    "        if len(file) != 0:\n",
    "            #shutil.copy2(img_t2w_, train_set+'img_3D/'+idx+'.nii.gz')\n",
    "            #shutil.copy2(img_cortex_, train_set+'CP_mask3D/'+idx+'.nii.gz')\n",
    "            shutil.copy2(img_cortex_v2_, train_set+'drawem9_multi_v4/mask3D_drawem9_multi_v4/'+idx+'.nii.gz')\n",
    "            #shutil.copy2(img_liq_, train_set+'total_mask3D/'+idx+'.nii.gz')\n",
    "            count += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To plot the results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 1.]\n",
      "1.0\n",
      "0.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 8 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_set = '/Users/meritxell/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/training_set/'\n",
    "total_path = train_set+'total_mask3D/CC00218AN12_74801.nii.gz'\n",
    "mask_path = train_set+'CP_mask3D/CC00218AN12_74801.nii.gz'\n",
    "image_path = train_set+'img_3D/CC00218AN12_74801.nii.gz'\n",
    "mask_path_v2 = train_set+'drawem9_2/CP_mask3D_drawem9/CC00218AN12_74801.nii.gz'\n",
    "mask_path_23 = train_set+'drawem9_23/mask3D_drawem9_23/CC00218AN12_74801.nii.gz'\n",
    "mask_path_1 = train_set+'drawem9_1/mask3D_drawem9_1/CC00218AN12_74801.nii.gz'\n",
    "mask_path_3 = train_set+'drawem9_3/mask3D_drawem9_3/CC00218AN12_74801.nii.gz'\n",
    "mask_path_4 = train_set+'drawem9_4/mask3D_drawem9_4/CC00218AN12_74801.nii.gz'\n",
    "mask_path_5 = train_set+'drawem9_5/mask3D_drawem9_5/CC00218AN12_74801.nii.gz'\n",
    "mask_path_6 = train_set+'drawem9_6/mask3D_drawem9_6/CC00218AN12_74801.nii.gz'\n",
    "mask_path_7 = train_set+'drawem9_7/mask3D_drawem9_7/CC00218AN12_74801.nii.gz'\n",
    "mask_path_8 = train_set+'drawem9_15/mask3D_drawem9_15/CC00218AN12_74801.nii.gz'\n",
    "mask_path_all_5 = train_set+'drawem9_all_5/mask3D_drawem9_all_5/CC00218AN12_74801.nii.gz'\n",
    "mask_path_all_8 = train_set+'drawem9_all_8/mask3D_drawem9_all_8/CC00218AN12_74801.nii.gz'\n",
    "img = nib.load(image_path).get_fdata()\n",
    "mask = nib.load(mask_path).get_fdata()\n",
    "mask_t = nib.load(total_path).get_fdata()\n",
    "mask_v2 = nib.load(mask_path_v2).get_fdata()\n",
    "mask_23 = nib.load(mask_path_23).get_fdata()\n",
    "mask_1 = nib.load(mask_path_1).get_fdata()\n",
    "mask_3 = nib.load(mask_path_3).get_fdata()\n",
    "mask_4 = nib.load(mask_path_4).get_fdata()\n",
    "mask_5 = nib.load(mask_path_5).get_fdata()\n",
    "mask_6 = nib.load(mask_path_6).get_fdata()\n",
    "mask_7 = nib.load(mask_path_7).get_fdata()\n",
    "mask_8 = nib.load(mask_path_8).get_fdata()\n",
    "mask_all_5 = nib.load(mask_path_all_5).get_fdata()\n",
    "mask_all_8 = nib.load(mask_path_all_8).get_fdata()\n",
    "\n",
    "print(np.unique(mask_3))\n",
    "\n",
    "print(np.max(mask_t))\n",
    "print(np.min(mask_t))\n",
    "\n",
    "plt.figure(figsize=(15, 10))\n",
    "plt.subplot(241)\n",
    "plt.title('MRI image')\n",
    "plt.axis('off')\n",
    "plt.imshow(img[:,:,100], cmap='gray')\n",
    "plt.subplot(242)\n",
    "plt.title('drawem9 Mask (2)')\n",
    "plt.imshow(mask_v2[:,:,100], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.subplot(243)\n",
    "plt.title('drawem9 Mask (3)')\n",
    "plt.axis('off')\n",
    "plt.imshow(mask_3[:,:,100], cmap='gray')\n",
    "plt.subplot(244)\n",
    "plt.title('drawem9 Mask (4)')\n",
    "plt.imshow(mask_4[:,:,100], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.subplot(245)\n",
    "plt.title('drawem9 Mask (5)')\n",
    "plt.imshow(mask_5[:,:,100], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.subplot(246)\n",
    "plt.title('drawem9 Mask (6)')\n",
    "plt.imshow(mask_6[:,:,150], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.subplot(247)\n",
    "plt.title('drawem9 Mask (7)')\n",
    "plt.imshow(mask_7[:,:,100], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.subplot(248)\n",
    "plt.title('drawem9 Mask (8)')\n",
    "plt.imshow(mask_8[:,:,100], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_set = '/Users/meritxell/Desktop/UPC/DD/Final_Project/Data_analysis/dHCP/training_set/'\n",
    "total_path = train_set+'total_mask3D/CC00218AN12_74801.nii.gz'\n",
    "mask_path = train_set+'CP_mask3D/CC00218AN12_74801.nii.gz'\n",
    "image_path = train_set+'img_3D/CC00218AN12_74801.nii.gz'\n",
    "mask_path_all_5 = train_set+'drawem9_multi_v4/mask3D_drawem9_multi_v4/CC00218AN12_74801.nii.gz'\n",
    "img = nib.load(image_path).get_fdata()\n",
    "mask = nib.load(mask_path).get_fdata()\n",
    "mask_all_5 = nib.load(mask_path_all_5).get_fdata()\n",
    "\n",
    "mask_all_5_1= np.zeros((mask_all_5.shape))\n",
    "mask_all_5_1 = np.where(mask_all_5 == 1, 1, np.nan)\n",
    "mask_all_5_2 = np.zeros((mask_all_5.shape))\n",
    "mask_all_5_2 = np.where(mask_all_5 == 2, 1, np.nan)\n",
    "mask_all_5_3 = np.zeros((mask_all_5.shape))\n",
    "mask_all_5_3 = np.where(mask_all_5 == 3, 1, np.nan)\n",
    "mask_all_5_4 = np.zeros((mask_all_5.shape))\n",
    "mask_all_5_4 = np.where(mask_all_5 == 4, 1, np.nan)\n",
    "\n",
    "plt.figure(figsize=(15, 10))\n",
    "plt.subplot(151)\n",
    "plt.title('MRI image')\n",
    "plt.axis('off')\n",
    "plt.imshow(img[:,:,100], cmap='gray')\n",
    "plt.subplot(152)\n",
    "plt.title('drawem9 Mask (1+5)')\n",
    "plt.axis('off')\n",
    "plt.imshow(mask_all_5_1[:,:,100], cmap='gray')\n",
    "plt.subplot(153)\n",
    "plt.title('drawem9 Mask (2)')\n",
    "plt.imshow(mask_all_5_2[:,:,100], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.subplot(154)\n",
    "plt.title('drawem9 Mask (3)')\n",
    "plt.imshow(mask_all_5_3[:,:,100], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.subplot(155)\n",
    "plt.title('drawem9 Mask (4)')\n",
    "plt.imshow(mask_all_5_4[:,:,100], cmap='gray')\n",
    "plt.axis('off')\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}