Copia_de_Transformers_medicina_extraccion.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copia de Transformers_medicina_extraccion_2.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPueTx/01bx8dAdu3hawE1x",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Morilla123/Curso/blob/main/Copia_de_Transformers_medicina_extraccion.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CPH7fdOtZHYz"
      },
      "source": [
        "## Extraccion de informacion de reportes medicos\n",
        "\n",
        "El objetivo de este proyecto es extraer desde un reporte medico que contiene texto libre, informacion del paciente, informacion de la medicacion del paciente, y una imagen de un radiografia de torax, lo siguiente:\n",
        "\n",
        "- Texto del reporte de-identificado (sin informacion personal del paciente)\n",
        "- Medicamentos con los que es tratado el paciente\n",
        "- Prediccion de si en la radiografia es normal o presenta neumonia\n",
        "\n",
        "Documento a analizar: [Reporte](https://drive.google.com/file/d/15zA5XDQfrVIYZk0gHGLxQqt_up7-NjuP/view?usp=sharing)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7xGbzCOpZNis"
      },
      "source": [
        "*1*) De-identificar texto (quitar informacion personal)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gmq6uhQCZYNB"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zbgoqD3EZQGR"
      },
      "source": [
        "2) Obtener medicamentos del paciente"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_yJC1aeSZc4G"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GknBcOy1ZP42"
      },
      "source": [
        "3) Predecir si la radiografia corresponde a un paciente normal o con neumonia"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8q1YujdtZBuM"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GEFQ3mjY5Qt0"
      },
      "source": [
        "## Anonimizacion (nombres)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1vFnWc-N5QGj"
      },
      "source": [
        "%%capture\n",
        "! pip install transformers datasets\n",
        "from transformers import AutoTokenizer, TFAutoModelForQuestionAnswering\n",
        "import tensorflow as tf\n",
        "!pip install spacy\n",
        "from spacy import displacy\n",
        "import spacy"
      ],
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i8S8nYc57Vnx"
      },
      "source": [
        "text = \"Name: Matias Molinas Birth date: 10/03/1918 (54) 342- 5678906 Significant Medical History Mr. Matias Molinas is a 52 y/o accountant with hypercholesterolemia and polycythemia vera who has been in relatively good health. Diagnostic: Patient needs to take 50 mg of ibuprofen. Chest X-Ray:\"\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "48XYhrJX5w2h"
      },
      "source": [
        "nlp = spacy.load(\"en_core_web_sm\")"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "tw-iT7oj5n7E",
        "outputId": "7f5a07c7-57da-4d83-c1be-3fab1b2a52a1"
      },
      "source": [
        "doc = nlp(text)\n",
        "displacy.render(doc, style='ent')"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'<div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">Name: \\n<mark class=\"entity\" style=\"background: #aa9cfc; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\\n    Matias Molinas Birth\\n    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; text-transform: uppercase; vertical-align: middle; margin-left: 0.5rem\">PERSON</span>\\n</mark>\\n date: \\n<mark class=\"entity\" style=\"background: #bfe1d9; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\\n    10/03/1918\\n    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; text-transform: uppercase; vertical-align: middle; margin-left: 0.5rem\">DATE</span>\\n</mark>\\n (\\n<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\\n    54\\n    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; text-transform: uppercase; vertical-align: middle; margin-left: 0.5rem\">CARDINAL</span>\\n</mark>\\n) 342- \\n<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\\n    5678906\\n    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; text-transform: uppercase; vertical-align: middle; margin-left: 0.5rem\">CARDINAL</span>\\n</mark>\\n Significant Medical History Mr. \\n<mark class=\"entity\" style=\"background: #aa9cfc; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\\n    Matias Molinas\\n    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; text-transform: uppercase; vertical-align: middle; margin-left: 0.5rem\">PERSON</span>\\n</mark>\\n is a \\n<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\\n    52 y\\n    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; text-transform: uppercase; vertical-align: middle; margin-left: 0.5rem\">QUANTITY</span>\\n</mark>\\n/o accountant with hypercholesterolemia and polycythemia vera who has been in relatively good health. Diagnostic: Patient needs to take \\n<mark class=\"entity\" style=\"background: #bfe1d9; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\\n    50 mg\\n    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; text-transform: uppercase; vertical-align: middle; margin-left: 0.5rem\">TIME</span>\\n</mark>\\n of ibuprofen. \\n<mark class=\"entity\" style=\"background: #aa9cfc; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\\n    Chest X-Ray\\n    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; text-transform: uppercase; vertical-align: middle; margin-left: 0.5rem\">PERSON</span>\\n</mark>\\n:</div>'"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5FaorFYX84bp",
        "outputId": "ca14fb95-244a-48f0-b342-fe815a308810"
      },
      "source": [
        "for ent in doc.ents:\n",
        "    print(ent.text, ent.label_)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Matias Molinas Birth PERSON\n",
            "10/03/1918 DATE\n",
            "54 CARDINAL\n",
            "5678906 CARDINAL\n",
            "Matias Molinas PERSON\n",
            "52 y QUANTITY\n",
            "50 mg TIME\n",
            "Chest X-Ray PERSON\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N8w4E9SF9bvo"
      },
      "source": [
        "# The start_char of each entity PER\n",
        "per_starts = [ent.start_char for ent in doc.ents if ent.label_ == 'PERSON']\n",
        "\n",
        "# The end_char of each entity PER\n",
        "per_ends = [ent.end_char for ent in doc.ents if ent.label_ == 'PERSON']\n",
        "\n",
        "# print(f'per_start: {per_starts}\\nper_ends: {per_ends}')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8XyAORDg9MEC"
      },
      "source": [
        "# colaboration https://github.com/mnsosa/ner/blob/main/ner_spanish_ialab.ipynb\n",
        "# Anonymization\n",
        "text_anon = doc.text[0:per_starts[0]]\n",
        "text_anon = text_anon + '<NOMBRE>'\n",
        "\n",
        "per = 1\n",
        "while per < len(per_starts):\n",
        "    text_anon = text_anon + doc.text[per_ends[per-1]:per_starts[per]] + '<NOMBRE> '\n",
        "    per = per + 1"
      ],
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "k45sZWS99qWc",
        "outputId": "d41239ee-31fd-4a82-e898-f39d259bedf6"
      },
      "source": [
        "text_anon"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'Name: <NOMBRE> date: 10/03/1918 (54) 342- 5678906 Significant Medical History Mr. <NOMBRE>  is a 52 y/o accountant with hypercholesterolemia and polycythemia vera who has been in relatively good health. Diagnostic: Patient needs to take 50 mg of ibuprofen. <NOMBRE> '"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B3kk2yboZX-e"
      },
      "source": [
        "## Extract list of drugs and disease\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Vw1zusXjZ6Uw"
      },
      "source": [
        "Import librerias\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uifRQEF-Z5Eg"
      },
      "source": [
        "%%capture\n",
        "import requests"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1pvZ2V63ZhGM"
      },
      "source": [
        "def Use_API(text, url=\"https://bern.korea.ac.kr/plain\"):          # API = https://bern.korea.ac.kr/#\n",
        "  return requests.post(url, data={'sample_text': text}).json()"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "m0CU4HtsZw4f"
      },
      "source": [
        "def Extraxtion_data(text,out,type):\n",
        "\n",
        "  denotations=out['denotations']\n",
        "  relavent_terms=[]\n",
        "  \n",
        "  if type=='drug':\n",
        "    drug_terms=[]\n",
        "    for i in denotations:\n",
        "      if i['obj']== 'drug':\n",
        "        drug_terms.append((i['span']['begin'],i['span']['end']))\n",
        "   \n",
        "    for i in drug_terms:\n",
        "      start=i[0]\n",
        "      end=i[1]\n",
        "      relavent_terms.append(text[start:end])\n",
        "\n",
        "  if type=='disease':\n",
        "    species_terms=[]\n",
        "    for i in denotations:\n",
        "      if i['obj']== 'disease':\n",
        "        species_terms.append((i['span']['begin'],i['span']['end']))\n",
        "    \n",
        "    for i in species_terms:\n",
        "      start=i[0]\n",
        "      end=i[1]\n",
        "      relavent_terms.append(text[start:end])\n",
        "\n",
        "\n",
        "  # if type=='species':\n",
        "  #   species_terms=[]\n",
        "  #   for i in denotations:\n",
        "  #     if i['obj']== 'species':\n",
        "  #       species_terms.append((i['span']['begin'],i['span']['end']))\n",
        "    \n",
        "  #   for i in species_terms:\n",
        "  #     start=i[0]\n",
        "  #     end=i[1]\n",
        "  #     relavent_terms.append(text[start:end])\n",
        "\n",
        "  return(relavent_terms)"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H5z8GQnnbDc0"
      },
      "source": [
        "# text = \"X-rays were negative and physical assessment determined soft tissue damage to the lateral aspect of her ankle. She was initially treated with ice, an ace wrap, crutches and mild pain medications (Tylenol with codeine)\"\n",
        "text = \"Name: Matias Molinas Birth date: 10/03/1918 (54) 342- 5678906 Significant Medical History Mr. Matias Molinas is a 52 y/o accountant with hypercholesterolemia and polycythemia vera who has been in relatively good health. Diagnostic: Patient needs to take 50 mg of ibuprofen. Chest X-Ray:\""
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XiVj6O3ubEyj"
      },
      "source": [
        "out=Use_API(text)\n",
        "# print(out)\n",
        "# print(Extraxtion_data(text,out,'drug'))\n",
        "List_Drugs = Extraxtion_data(text,out,'drug')\n",
        "# print(Extraxtion_data(text,out,'disease'))\n",
        "List_Disease = Extraxtion_data(text,out,'disease')"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ntJMgy77eu9t",
        "outputId": "d79dce2c-b059-4fe9-ea23-87049371fcc0"
      },
      "source": [
        "print(List_Drugs, List_Disease)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['ibuprofen'] ['hypercholesterolemia', 'polycythemia vera']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aYAEoKMz0SID"
      },
      "source": [
        "## image extract (check input path)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MR-KXSgZ0RCx"
      },
      "source": [
        "%%capture\n",
        "!pip install PyMuPDF\n",
        "import fitz"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hxuZE90z1T5K",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3e66b942-b9ca-4da5-c97f-987070647982"
      },
      "source": [
        "file_path = input(\"Enter the PDF file path\")\n",
        "pdf_file = fitz.open(file_path)\n",
        "\n",
        "\n",
        "\n",
        "#Reading the location where to save the file\n",
        "location = input(\"Enter the location to save: \")\n",
        "\n",
        "\n",
        "#finding number of pages in the pdf\n",
        "number_of_pages = len(pdf_file)\n",
        "\n",
        "#iterating through each page in the pdf\n",
        "for current_page_index in range(number_of_pages):\n",
        "  #iterating through each image in every page of PDF\n",
        "  for img_index,img in enumerate(pdf_file.getPageImageList(current_page_index)):\n",
        "        xref = img[0]\n",
        "        image = fitz.Pixmap(pdf_file, xref)\n",
        "        #if it is a is GRAY or RGB image\n",
        "        if image.n < 5:        \n",
        "            image.writePNG(\"{}/image{}-{}.png\".format(location,current_page_index, img_index))\n",
        "        #if it is CMYK: convert to RGB first\n",
        "        else:                \n",
        "            new_image = fitz.Pixmap(fitz.csRGB, image)\n",
        "            new_image.writePNG(\"{}/image{}-{}.png\".foramt(location,current_page_index, img_index))\n",
        "            "
      ],
      "execution_count": 16,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Enter the PDF file path/content/Report.pdf\n",
            "Enter the location to save: /content/\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Deprecation: 'getPageImageList' removed from class 'Document' after v1.19.0 - use 'get_page_images'.\n",
            "Deprecation: 'writePNG' removed from class 'Pixmap' after v1.19.0 - use 'save'.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BK6UPCMJ-PIQ"
      },
      "source": [
        "## Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K6Vs2uDn-UOY"
      },
      "source": [
        "%%capture\n",
        "!pip install kaggle --upgrade"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4MiUtVk8KdQO"
      },
      "source": [
        "! mkdir ~/.kaggle # simplemente creo una carpeta en colab pero en la raiz"
      ],
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BF9JxOOeK1Gj"
      },
      "source": [
        "! cp kaggle.json ~/.kaggle/ # mueve el kaggle jonson para la carpeta que acabamos de crear"
      ],
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XkeF4SDELNnt"
      },
      "source": [
        "! chmod 600 ~/.kaggle/kaggle.json # es como el permiso para que funque"
      ],
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EROLUbsKLRfQ"
      },
      "source": [
        "%%capture\n",
        "! kaggle datasets download paultimothymooney/chest-xray-pneumonia"
      ],
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O3IEVTR5PMSd"
      },
      "source": [
        "! unzip chest-xray-pneumonia.zip"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H-fIjJ3ARPT5"
      },
      "source": [
        "%%capture\n",
        "!pip install -U tensorflow-addons"
      ],
      "execution_count": 30,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nLpRxWyIYFJz"
      },
      "source": [
        "## Create Baseline"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p1cR6W3wRgXt"
      },
      "source": [
        "%%capture\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "import tensorflow_addons as tfa\n",
        "\n",
        "import numpy as np # linear algebra\n",
        "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
        "import cv2 # import cv2\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "from PIL import Image\n",
        "import random"
      ],
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0xIx20zjR3Vd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e5f7f517-c82a-4aad-ae64-5d01ecf726f4"
      },
      "source": [
        "import os\n",
        "# print(os.listdir(\"/content/chest_xray\"))"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['__MACOSX', 'train', 'val', 'test', 'chest_xray']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cYJOcDqZNAax",
        "outputId": "e1b78fbe-2169-4104-db55-33cc66294f48"
      },
      "source": [
        "DIR = os.listdir('/content/chest_xray')\n",
        "print(DIR)"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['__MACOSX', 'train', 'val', 'test', 'chest_xray']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LUWpfXo2NPQH"
      },
      "source": [
        "DIR = os.listdir('/content/chest_xray')\n",
        "\n",
        "train_folder = '/content/chest_xray/train'\n",
        "test_folder = '/content/chest_xray/test'\n",
        "val_folder = '/content/chest_xray/val'"
      ],
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xzI5Eq9HOc10"
      },
      "source": [
        "IMG_SIZE = 50 # resize image, see  u  later             -----------------------------------------------------------------------"
      ],
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w93ib43SNPEo"
      },
      "source": [
        "labels = [\"NORMAL\", \"PNEUMONIA\"] # each folder has two sub folder name \"PNEUMONIA\", \"NORMAL\"\n",
        "\n",
        "def get_data_train(data_dir):\n",
        "    data = []\n",
        "    for label in labels:\n",
        "        path = os.path.join(data_dir, label)\n",
        "        class_num = labels.index(label)\n",
        "        for img in os.listdir(path):\n",
        "            try:\n",
        "                img_array = cv2.imread(os.path.join(path, img), cv2.IMREAD_GRAYSCALE)\n",
        "                new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))\n",
        "                data.append([new_array, class_num])\n",
        "            except Exception as e:\n",
        "                print(e)\n",
        "    return np.array(data)"
      ],
      "execution_count": 43,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0XbnxVTeNO-u"
      },
      "source": [
        "train = get_data_train(train_folder)\n",
        "test = get_data_train(test_folder)\n",
        "val = get_data_train(val_folder)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2H3cyJdRRno5"
      },
      "source": [
        "l = []                    <--------------- make the variable understandable"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 360
        },
        "id": "TSGlFwcjNO7-",
        "outputId": "c3cefdb4-25f2-46ee-845d-a1f172f2d1f7"
      },
      "source": [
        "l = []\n",
        "for i in train:\n",
        "    if(i[1] == 0):\n",
        "        l.append(\"Normal\")\n",
        "    else:\n",
        "        l.append(\"Pneumonia\")\n",
        "        \n",
        "sns.countplot(l)"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variable as a keyword arg: x. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
            "  FutureWarning\n",
            "\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f8ff8bd0390>"
            ]
          },
          "metadata": {},
          "execution_count": 47
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Zw_BPKDPWDMD"
      },
      "source": [
        "X_train = []\n",
        "y_train = []\n",
        "\n",
        "X_val = []\n",
        "y_val = []\n",
        "\n",
        "X_test = []\n",
        "y_test = []\n",
        "\n",
        "for feature, label in train:\n",
        "    X_train.append(feature)\n",
        "    y_train.append(label)\n",
        "\n",
        "for feature, label in test:\n",
        "    X_test.append(feature)\n",
        "    y_test.append(label)\n",
        "    \n",
        "for feature, label in val:\n",
        "    X_val.append(feature)\n",
        "    y_val.append(label)"
      ],
      "execution_count": 48,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VGgoQo12NOxn"
      },
      "source": [
        "X_train = np.array(X_train) / 255\n",
        "X_val = np.array(X_val) / 255\n",
        "X_test = np.array(X_test) / 255 # normalizacion"
      ],
      "execution_count": 49,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UAtzXw0jWM1d",
        "outputId": "1cdf40e6-3ffe-4097-c304-43e26b2314e0"
      },
      "source": [
        "# print(X_test.shape, X_val.shape, X_test.shape )"
      ],
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(624, 50, 50) (16, 50, 50) (624, 50, 50)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F-eV12ODXoC2"
      },
      "source": [
        "X_train = X_train.reshape(-1, IMG_SIZE, IMG_SIZE, 1)\n",
        "y_train = np.array(y_train)\n",
        "\n",
        "X_val = X_val.reshape(-1, IMG_SIZE, IMG_SIZE, 1)\n",
        "y_val = np.array(y_val)\n",
        "\n",
        "X_test = X_test.reshape(-1, IMG_SIZE, IMG_SIZE, 1)\n",
        "y_test = np.array(y_test)"
      ],
      "execution_count": 53,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lu-_ayERXs6d",
        "outputId": "9cdd2add-1b1f-42e2-8dd1-ca4ef8406e78"
      },
      "source": [
        "# print(X_test.shape, X_val.shape, X_test.shape )"
      ],
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(624, 50, 50, 1) (16, 50, 50, 1) (624, 50, 50, 1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_G5LMT8SX38n"
      },
      "source": [
        "%%capture\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.layers import Flatten, Conv2D, Activation, Dense, Dropout, MaxPooling2D\n",
        "from tensorflow.keras.models import Sequential"
      ],
      "execution_count": 55,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PWZZSkJnYYyd"
      },
      "source": [
        "Create vanillia model "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GHxYFfQ4YcMx"
      },
      "source": [
        "model = Sequential()\n",
        "\n",
        "model.add(Conv2D(32, (3, 3), padding=\"same\", input_shape=X_train.shape[1:]))\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(2, 2))\n",
        "model.add(Dropout(0.2))\n",
        "\n",
        "model.add(Conv2D(64, (3, 3), padding=\"same\"))\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(2, 2))\n",
        "model.add(Dropout(0.2))\n",
        "\n",
        "model.add(Conv2D(128, (3, 3), padding=\"same\"))\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(2, 2))\n",
        "model.add(Dropout(0.2))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(128, activation=\"relu\"))\n",
        "\n",
        "model.add(Dense(1))\n",
        "model.add(Activation(\"sigmoid\"))"
      ],
      "execution_count": 56,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XobwekznYnoP"
      },
      "source": [
        "%%capture\n",
        "model.compile(loss=\"binary_crossentropy\", optimizer=\"adam\", metrics=[\"accuracy\"])\n",
        "history = model.fit(X_train, y_train, epochs=25, validation_data=(X_val, y_val), shuffle=True)\n",
        "scores = model.evaluate(X_test, y_test)"
      ],
      "execution_count": 74,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q4_tGfO2b4Cc"
      },
      "source": [
        "model.save(\"cnn_baseline.model\")"
      ],
      "execution_count": 66,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Tpa-m_SrbKgG",
        "outputId": "6ee0f15c-642d-44ce-d456-3f346c58f787"
      },
      "source": [
        "# scores\n",
        "print(\"Test loss {}\".format(scores[0]))\n",
        "print(\"Test accuracy {}\".format(scores[1]))"
      ],
      "execution_count": 62,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test loss 3.2666449546813965\n",
            "Test accuracy 0.7435897588729858\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 513
        },
        "id": "ICeiii6tbOOB",
        "outputId": "9919c6ac-63e3-444a-dd8b-e7b9237f54aa"
      },
      "source": [
        "accuracy = history.history['accuracy']\n",
        "val_accuracy = history.history['val_accuracy']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "epochs = range(len(accuracy))\n",
        "\n",
        "plt.plot(epochs, accuracy, \"b\", label=\"trainning accuracy\")\n",
        "plt.plot(epochs, val_accuracy, \"r\", label=\"validation accuracy\")\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "plt.plot(epochs, loss, \"b\", label=\"trainning loss\")\n",
        "plt.plot(epochs, val_loss, \"r\", label=\"validation loss\")\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 63,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I2sb43qubvaA",
        "outputId": "4b02bfc2-a838-4174-dfd2-ba4dde37dee6"
      },
      "source": [
        "predict_x=model.predict(X_test) \n",
        "prediction=np.argmax(predict_x,axis=1)\n",
        "prediction[:15]"
      ],
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])"
            ]
          },
          "metadata": {},
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BpstZi5cbe8t"
      },
      "source": [
        "correct = np.nonzero(prediction == y_test)[0]\n",
        "incorrect = np.nonzero(prediction != y_test)[0]"
      ],
      "execution_count": 77,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "rP2HzPY-d7Sr",
        "outputId": "7b953a8e-fa01-4d24-db0e-8525396109e1"
      },
      "source": [
        "# correc\n",
        "j = 0\n",
        "for i in correct[:6]:\n",
        "    plt.subplot(3,2,j+1)\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "    plt.imshow(X_test[i].reshape(50,50), cmap=\"gray\", interpolation='none')\n",
        "    plt.title(\"Predicted Class {},Actual Class {}\".format(prediction[i], y_test[i]))\n",
        "    plt.xlabel(labels[prediction[i]])\n",
        "    plt.tight_layout()\n",
        "    j += 1"
      ],
      "execution_count": 78,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 6 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "5MhV6gU3eioP",
        "outputId": "fa127971-e8eb-4563-c38c-b86561282ec6"
      },
      "source": [
        "# incorrect\n",
        "j = 0\n",
        "for i in incorrect[:6]:\n",
        "    plt.subplot(3,2,j+1)\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "    plt.imshow(X_test[i].reshape(50,50), cmap=\"gray\", interpolation='none')\n",
        "    plt.title(\"Predicted Class {},Actual Class {}\".format(prediction[i], y_test[i]))\n",
        "    plt.xlabel(labels[prediction[i]])\n",
        "    plt.tight_layout()\n",
        "    j += 1"
      ],
      "execution_count": 79,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 6 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h24FFhnKcg6O"
      },
      "source": [
        "labels = [\"NORMAL\", \"PNEUMONIA\"]\n",
        "def prepare(filepath):\n",
        "    img_array = cv2.imread(filepath, cv2.IMREAD_GRAYSCALE)\n",
        "    new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))\n",
        "    return new_array.reshape(-1, IMG_SIZE, IMG_SIZE, 1)\n",
        "\n",
        "model = tf.keras.models.load_model(\"cnn_baseline.model\") # load model"
      ],
      "execution_count": 75,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ouSTZpcFe1YL"
      },
      "source": [
        "labels = [\"NORMAL\", \"PNEUMONIA\"]\n",
        "def prepare(filepath):\n",
        "    img_array = cv2.imread(filepath, cv2.IMREAD_GRAYSCALE)\n",
        "    new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))\n",
        "    return new_array.reshape(-1, IMG_SIZE, IMG_SIZE, 1)\n",
        "\n",
        "model = tf.keras.models.load_model(\"cnn_baseline.model\") # load model"
      ],
      "execution_count": 80,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hjEamiQfdidZ"
      },
      "source": [
        "## ViT"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0TgXuMcndgm7"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}