tumor-classifier.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "7bb66c1a",
   "metadata": {},
   "source": [
    "# Brain Tumor Segmentation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0b3497c",
   "metadata": {},
   "source": [
    "## (Un)Healthy Brain Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5603a606",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import cv2\n",
    "import urllib\n",
    "from PIL import Image\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn.model_selection import train_test_split\n",
    "import albumentations\n",
    "import time\n",
    "import argparse\n",
    "from tqdm import tqdm\n",
    "import warnings\n",
    "import sklearn as sk\n",
    "import itertools\n",
    "import zlib\n",
    "\n",
    "# Pytorch libraries and subsidiaries\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import torch.cuda as cuda\n",
    "import torchvision as tv\n",
    "from torch.utils.data import TensorDataset, DataLoader, Dataset\n",
    "from torchvision import transforms as T\n",
    "from torchsummary import summary # allows for keras-like summaries of models in pytorch\n",
    "from torchvision.utils import make_grid"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2aba66f1",
   "metadata": {},
   "source": [
    "### Check CUDA Capability and Set Device"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "88a72b3d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running on NVIDIA GeForce GTX 1070\n"
     ]
    }
   ],
   "source": [
    "if cuda.is_available():\n",
    "    device = torch.device('cuda:0')\n",
    "    print('Running on ' + cuda.get_device_name(device))\n",
    "else:\n",
    "    device = torch.device('cpu')\n",
    "    print('Running on the CPU')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f74051ea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<PIL.PngImagePlugin.PngImageFile image mode=L size=240x300 at 0x221FCCC7A60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(300, 240, 3)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Download an example image\n",
    "path = ('brain-tumor-dataset/Brain Tumor/Cancer (1).png')\n",
    "\n",
    "input_image = Image.open(path)\n",
    "\n",
    "display(input_image)\n",
    "cv2.imread(path).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20cf294e",
   "metadata": {},
   "source": [
    "### Create Pandas Dataframe of Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a88da1ce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "filepaths:  8117    labels:  8117\n",
      "1    5432\n",
      "0    2685\n",
      "Name: labels, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "tumor_dir = 'brain-tumor-dataset/Brain Tumor'\n",
    "healthy_dir = 'brain-tumor-dataset/Healthy'\n",
    "dirlist = [tumor_dir, healthy_dir]\n",
    "classes = [1, 0] # 1 = tumor, 0 = no tumor\n",
    "filepaths = []\n",
    "labels = []\n",
    "for d,c in zip(dirlist, classes):\n",
    "    flist = os.listdir(d)\n",
    "    for f in flist:\n",
    "        fpath = os.path.join (d,f)\n",
    "        filepaths.append(fpath)\n",
    "        labels.append(c)\n",
    "print ('filepaths: ', len(filepaths), '   labels: ', len(labels))\n",
    "\n",
    "Fseries = pd.Series(filepaths, name='file_paths')\n",
    "Lseries = pd.Series(labels, name='labels')\n",
    "df = pd.concat([Fseries,Lseries], axis=1, names=['file_paths', 'labels'])\n",
    "df = pd.DataFrame(np.array(df).reshape(8117,2), columns=['file_paths', 'labels'])\n",
    "print(df['labels'].value_counts())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "932adece",
   "metadata": {},
   "source": [
    "### Dataset Splitting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "72541c27",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Set:\n",
      "1    4885\n",
      "0    2420\n",
      "Name: labels, dtype: int64\n",
      "Validation Set:\n",
      "1    278\n",
      "0    128\n",
      "Name: labels, dtype: int64\n",
      "Test Set:\n",
      "1    269\n",
      "0    137\n",
      "Name: labels, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "train_df, test_df = train_test_split(df, train_size=0.90, random_state=0)\n",
    "test_df, valid_df = train_test_split(test_df, train_size=0.5, random_state=0)\n",
    "\n",
    "print('Training Set:')\n",
    "print(train_df.labels.value_counts())\n",
    "print('Validation Set:')\n",
    "print(valid_df.labels.value_counts())\n",
    "print('Test Set:')\n",
    "print(test_df.labels.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "eb2de3d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = train_df.file_paths.values\n",
    "y_train = train_df.labels.values\n",
    "X_valid = valid_df.file_paths.values\n",
    "y_valid = valid_df.labels.values\n",
    "X_test = test_df.file_paths.values\n",
    "y_test = test_df.labels.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7ff67e47",
   "metadata": {},
   "outputs": [],
   "source": [
    "class BrainMRIDataset(Dataset):\n",
    "    def __init__(self, X, y):\n",
    "        self.X = X \n",
    "        self.y = y\n",
    "    \n",
    "    transform = T.Compose([\n",
    "        T.Resize((224, 224)),\n",
    "        T.ToTensor()])\n",
    "\n",
    "    def __len__(self):\n",
    "        return (len(self.X))\n",
    "    \n",
    "    def __getitem__(self, i):\n",
    "        image = Image.open(self.X[i])\n",
    "        image = image.convert('RGB')\n",
    "        image = self.transform(image)\n",
    "        image = np.transpose(image, (0, 2, 1))\n",
    "        label = self.y[i]\n",
    "        return torch.tensor(image, dtype=torch.float), torch.tensor(label, dtype=torch.long)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "08b11abf",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = BrainMRIDataset(X_train, y_train)\n",
    "valid_data = BrainMRIDataset(X_valid, y_valid)\n",
    "test_data = BrainMRIDataset(X_test, y_test)\n",
    "\n",
    "trainloader = DataLoader(train_data, batch_size=32, shuffle=True)\n",
    "validloader = DataLoader(valid_data, batch_size=1, shuffle=False)\n",
    "testloader = DataLoader(test_data, batch_size=1, shuffle=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "594c601f",
   "metadata": {},
   "source": [
    "### ResNet Model Definition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "f5e9d8b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "def resnet_model(pretrained, requires_grad):\n",
    "    resnet_model = tv.models.resnet152(progress=True, pretrained=pretrained)\n",
    "    # freeze hidden layers\n",
    "    if requires_grad == False:\n",
    "        for param in resnet_model.parameters():\n",
    "            param.requires_grad = False\n",
    "    # train the hidden layers\n",
    "    elif requires_grad == True:\n",
    "        for param in resnet_model.parameters():\n",
    "            param.requires_grad = True\n",
    "    # make the classification layer learnable\n",
    "    resnet_model.fc = nn.Linear(resnet_model.fc.in_features, 2)\n",
    "    return resnet_model\n",
    "    \n",
    "resnet_model = resnet_model(pretrained=True, requires_grad=False).to(device) # Frozen hidden layers model \n",
    "# print(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "701cf14c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# optimizer\n",
    "optimizer = optim.SGD(resnet_model.parameters(), lr=1e-3, momentum=0.85, weight_decay=1e-4)\n",
    "# loss function\n",
    "criterion = nn.CrossEntropyLoss().cuda() if torch.cuda.is_available() else nn.CrossEntropyLoss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "bbb17c23",
   "metadata": {},
   "outputs": [],
   "source": [
    "# validation function\n",
    "def validate(model, valid_dataloader):\n",
    "    print('Validating')\n",
    "    model.eval()\n",
    "    val_running_loss = 0.0\n",
    "    val_running_correct = 0\n",
    "    \n",
    "    with torch.no_grad():\n",
    "        for i, data in tqdm(enumerate(valid_dataloader), total=int(len(valid_data)/valid_dataloader.batch_size)):\n",
    "            data, target = data[0].to(device), data[1].to(device)\n",
    "            outputs = model(data)\n",
    "            loss = criterion(outputs, target)\n",
    "            \n",
    "            val_running_loss += loss.item()\n",
    "            _, preds = torch.max(outputs.data, 1)\n",
    "            val_running_correct += (preds == target).sum().item()\n",
    "        \n",
    "        val_loss = val_running_loss/len(valid_dataloader.dataset)\n",
    "        val_accuracy = 100. * val_running_correct/len(valid_dataloader.dataset)\n",
    "        print(f'Val Loss: {val_loss:.4f}, Val Acc: {val_accuracy:.2f}')\n",
    "        \n",
    "        return val_loss, val_accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "77dd4aaa",
   "metadata": {},
   "outputs": [],
   "source": [
    "# training function\n",
    "def fit(model, train_dataloader):\n",
    "    print('Training')\n",
    "    model.train()\n",
    "    train_running_loss = 0.0\n",
    "    train_running_correct = 0\n",
    "    \n",
    "    for i, data in tqdm(enumerate(train_dataloader), total=int(len(train_data)/train_dataloader.batch_size)):\n",
    "        data, target = data[0].to(device), data[1].to(device)\n",
    "        optimizer.zero_grad()\n",
    "        outputs = model(data)\n",
    "        loss = criterion(outputs, target)\n",
    "        train_running_loss += loss.item()\n",
    "        _, preds = torch.max(outputs.data, 1)\n",
    "        train_running_correct += (preds == target).sum().item()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        \n",
    "    train_loss = train_running_loss / len(train_dataloader.dataset)\n",
    "    train_accuracy = 100. * train_running_correct / len(train_dataloader.dataset)\n",
    "    \n",
    "    print(f\"Train Loss: {train_loss:.4f}, Train Acc: {train_accuracy:.2f}\")\n",
    "    \n",
    "    return train_loss, train_accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "cc268ab0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CUDA Enabled: True\n",
      "Epoch 1 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:27,  2.61it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 28.30it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0115, Train Acc: 83.61\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 29.74it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.2473, Val Acc: 91.13\n",
      "Epoch 2 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:24,  2.72it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 27.28it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0081, Train Acc: 90.42\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.65it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.2170, Val Acc: 92.12\n",
      "Epoch 3 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:26,  2.66it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 28.30it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0071, Train Acc: 91.42\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 29.95it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1957, Val Acc: 92.86\n",
      "Epoch 4 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:27,  2.62it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 28.17it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0068, Train Acc: 92.06\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.94it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1800, Val Acc: 93.60\n",
      "Epoch 5 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:25,  2.67it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:16, 24.90it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0064, Train Acc: 92.65\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 29.68it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1712, Val Acc: 93.60\n",
      "Epoch 6 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:25,  2.68it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:13, 28.85it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0059, Train Acc: 93.18\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 29.53it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1781, Val Acc: 93.10\n",
      "Epoch 7 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:26,  2.65it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 28.30it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0057, Train Acc: 93.43\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.83it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1630, Val Acc: 93.35\n",
      "Epoch 8 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:27,  2.62it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 28.44it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0057, Train Acc: 93.55\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 29.79it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1509, Val Acc: 94.83\n",
      "Epoch 9 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:24,  2.71it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 28.04it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0054, Train Acc: 93.73\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 31.02it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1469, Val Acc: 95.57\n",
      "Epoch 10 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:24,  2.72it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:13, 29.27it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0056, Train Acc: 93.20\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:12<00:00, 31.44it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1514, Val Acc: 94.09\n",
      "Epoch 11 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:23,  2.73it/s]                         \n",
      "  1%|          | 4/406 [00:00<00:12, 31.50it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0051, Train Acc: 94.10\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:12<00:00, 31.24it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1699, Val Acc: 92.61\n",
      "Epoch 12 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:24,  2.72it/s]                         \n",
      "  1%|          | 4/406 [00:00<00:12, 33.47it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0053, Train Acc: 93.87\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:12<00:00, 31.29it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1541, Val Acc: 93.10\n",
      "Epoch 13 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:23,  2.73it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 27.65it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0051, Train Acc: 94.20\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 31.13it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1472, Val Acc: 93.35\n",
      "Epoch 14 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:24,  2.72it/s]                         \n",
      "  1%|▏         | 6/406 [00:00<00:13, 29.11it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0049, Train Acc: 94.43\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:12<00:00, 31.24it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1408, Val Acc: 93.60\n",
      "Epoch 15 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:24,  2.70it/s]                         \n",
      "  1%|▏         | 6/406 [00:00<00:14, 28.48it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0049, Train Acc: 94.11\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.76it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1387, Val Acc: 94.33\n",
      "Epoch 16 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:25,  2.67it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:16, 25.10it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0048, Train Acc: 94.29\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.17it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1432, Val Acc: 94.33\n",
      "Epoch 17 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:27,  2.61it/s]                         \n",
      "  1%|          | 4/406 [00:00<00:12, 32.39it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0048, Train Acc: 94.69\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 29.44it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1285, Val Acc: 95.07\n",
      "Epoch 18 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:25,  2.68it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:15, 25.86it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0046, Train Acc: 94.85\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.38it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1260, Val Acc: 95.57\n",
      "Epoch 19 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:25,  2.67it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:15, 26.79it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0047, Train Acc: 94.58\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:14<00:00, 28.11it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1248, Val Acc: 94.83\n",
      "Epoch 20 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:24,  2.70it/s]                         \n",
      "  1%|          | 4/406 [00:00<00:12, 31.37it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0047, Train Acc: 94.54\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.43it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1469, Val Acc: 95.07\n",
      "Epoch 21 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:23,  2.75it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:13, 29.27it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0047, Train Acc: 94.28\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.07it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1323, Val Acc: 95.07\n",
      "Epoch 22 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:28,  2.59it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:15, 25.64it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0044, Train Acc: 95.02\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:14<00:00, 28.20it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1271, Val Acc: 95.32\n",
      "Epoch 23 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:31,  2.51it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:15, 26.55it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0045, Train Acc: 94.74\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:14<00:00, 28.10it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1258, Val Acc: 95.07\n",
      "Epoch 24 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:26,  2.66it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 27.78it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0043, Train Acc: 94.99\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.72it/s]\n",
      "  0%|          | 0/228 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1302, Val Acc: 95.07\n",
      "Epoch 25 of 25\n",
      "Training\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "229it [01:22,  2.76it/s]                         \n",
      "  1%|          | 3/406 [00:00<00:14, 27.03it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Loss: 0.0042, Train Acc: 95.22\n",
      "Validating\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 406/406 [00:13<00:00, 30.74it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Val Loss: 0.1276, Val Acc: 95.32\n",
      "41.301 minutes\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x540 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# %%capture\n",
    "# %%script false --no-raise-error\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "print('CUDA Enabled: ' + str(next(resnet_model.parameters()).is_cuda))\n",
    "\n",
    "train_loss , train_accuracy = [], []\n",
    "val_loss , val_accuracy = [], []\n",
    "start = time.time()\n",
    "\n",
    "for epoch in range(25):\n",
    "    print(f\"Epoch {epoch + 1} of 25\")\n",
    "    train_epoch_loss, train_epoch_accuracy = fit(resnet_model, trainloader)\n",
    "    val_epoch_loss, val_epoch_accuracy = validate(resnet_model, validloader)\n",
    "    train_loss.append(train_epoch_loss)\n",
    "    train_accuracy.append(train_epoch_accuracy)\n",
    "    val_loss.append(val_epoch_loss)\n",
    "    val_accuracy.append(val_epoch_accuracy)\n",
    "\n",
    "end = time.time()\n",
    "print(f\"{(end-start) / 60:.3f} minutes\")\n",
    "\n",
    "y = val_loss\n",
    "x = [ i for i in range(1, 26) ]\n",
    "fig = plt.figure(figsize=(10,7.5))\n",
    "ax = fig.add_subplot(111)\n",
    "ax.plot(x, y)\n",
    "_ = plt.xlabel(\"Epochs\")\n",
    "_ = plt.ylabel(\"Validation Loss\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "d9159a83",
   "metadata": {},
   "outputs": [],
   "source": [
    "# %%capture --no-stderr\n",
    "# %%script false --no-raise-error\n",
    "PATH = './resnet_brain_recent.pth'\n",
    "torch.save(resnet_model, PATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "6a988e7e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# if torch.cuda.is_available():\n",
    "#     resnet_model = torch.load('./resnet_brain_v2.pth', map_location='cuda:0')\n",
    "# else:\n",
    "#     resnet_model = torch.load('./resnet_brain_v2.pth', map_location=torch.device('cpu'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "51d3ca56",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Testing\n",
    "warnings.filterwarnings('ignore')\n",
    "correct = 0\n",
    "total = 0\n",
    "X_wrong = []\n",
    "if torch.cuda.is_available():\n",
    "    y_pred = torch.Tensor([]).cuda()\n",
    "else:\n",
    "    y_pred = torch.Tensor([])\n",
    "\n",
    "resnet_model.eval()\n",
    "\n",
    "with torch.no_grad():\n",
    "    for data in testloader:\n",
    "        images, labels = data\n",
    "        if torch.cuda.is_available():\n",
    "            images, labels = images.cuda(), labels.cuda()\n",
    "        outputs = resnet_model(images) # feed images into resnet and get predictions\n",
    "        y_pred = torch.cat((y_pred, outputs), dim=0)\n",
    "        _, predicted = torch.max(outputs.data, 1)\n",
    "        total += labels.size(0)\n",
    "        correct += (predicted == labels).sum().item()\n",
    "        if(predicted != labels):\n",
    "            X_wrong.append((images.cpu(), labels.cpu()))\n",
    "\n",
    "y_pred = np.uint8(np.argmax(y_pred.cpu(), axis=1))\n",
    "y_test = np.uint8(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "c32db061",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.93      0.92      0.92       137\n",
      "           1       0.96      0.96      0.96       269\n",
      "\n",
      "    accuracy                           0.95       406\n",
      "   macro avg       0.94      0.94      0.94       406\n",
      "weighted avg       0.95      0.95      0.95       406\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(sk.metrics.classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "dcea1984",
   "metadata": {},
   "outputs": [],
   "source": [
    "conf_matrix = sk.metrics.confusion_matrix(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "30167e06",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(cm, classes, normalize=False, title='Confusion matrix', cmap=plt.cm.Blues):\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    print(cm)\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt), horizontalalignment=\"center\", color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e44b20cb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[126  11]\n",
      " [ 10 259]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(conf_matrix, ['tumor', 'healthy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "f866334f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Misclassified: 21\n",
      "Predicted Label: Tumor\n",
      "True Label: Healthy\n",
      "False Positive\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x221a0ead340>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "plt.axis('off')\n",
    "wrong_index = 11\n",
    "print('Total Misclassified: ' + str(len(X_wrong)))\n",
    "if X_wrong[wrong_index][1].item() == 1:\n",
    "    print('Predicted Label: Healthy\\nTrue Label: Tumor\\nFalse Negative')\n",
    "else:\n",
    "    print('Predicted Label: Tumor\\nTrue Label: Healthy\\nFalse Positive')\n",
    "    \n",
    "plt.imshow(X_wrong[wrong_index][0].permute(3, 2, 0, 1).squeeze())"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "d6ecb163a55f96fb88d1210187b3ce0fa2f8272fd5184f376e8cea9f3c0de3eb"
  },
  "kernelspec": {
   "display_name": "Python 3.8.8 64-bit (conda)",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}