Metaanalysis using single subject maps.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This needs to be installed inside the notebook for compatibility with Kaggle\n",
    "!pip install git+git://github.com/neurostuff/nimare.git@608516ec3034e356326dfe70df5e9ed77efd2be8#egg=nimare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\me\\Miniconda3\\lib\\importlib\\_bootstrap.py:219: ImportWarning: can't resolve package from __spec__ or __package__, falling back on __name__ and __path__\n",
      "  return f(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import io\n",
    "import requests\n",
    "import pandas as pd\n",
    "from nilearn.image import resample_to_img, math_img\n",
    "from nilearn.datasets import load_mni152_template, load_mni152_brain_mask\n",
    "import nibabel as nb\n",
    "from gzip import GzipFile\n",
    "from nimare.meta.ibma import stouffers, weighted_stouffers\n",
    "from nilearn.masking import apply_mask\n",
    "from nilearn.plotting import plot_roi, plot_stat_map, plot_glass_brain\n",
    "\n",
    "template_nii = load_mni152_template()\n",
    "template_mask_nii = load_mni152_brain_mask()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This tutorial shows how to use NeuroVault API and NiMARE to perfroma a metaanalysis on a set of single subject maps"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Defining which images will be used in the metaanalysis\n",
    "We begin by creating a helper function to get metadata for all images that are part of a set of collections we are interested in"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_images_metadata(collection_ids):\n",
    "    images = []\n",
    "    for collection_id in collection_ids:\n",
    "        url = \"http://neurovault.org/api/collections/%d/images/?format=json\"%collection_id\n",
    "        while url:\n",
    "            r = requests.get(url)\n",
    "            d = r.json()\n",
    "            images += d['results']\n",
    "            url = d['next']\n",
    "    return pd.DataFrame(images)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this excercise we are coing to use only one collection but this list can be extended to multiple collections"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "collection_ids = [3235]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['BMI', 'SDS-BMI', 'add_date', 'age', 'analysis_level', 'bis11_score',\n",
       "       'bis_bas_score', 'brain_coverage', 'cognitive_contrast_cogatlas',\n",
       "       'cognitive_contrast_cogatlas_id', 'cognitive_paradigm_cogatlas',\n",
       "       'cognitive_paradigm_cogatlas_id', 'cognitive_paradigm_description_url',\n",
       "       'collection', 'collection_id', 'contrast_definition',\n",
       "       'contrast_definition_cogatlas', 'data_origin',\n",
       "       'days_since_menstruation', 'description', 'ethnicity', 'fat_percentage',\n",
       "       'figure', 'file', 'file_size', 'gender', 'handedness',\n",
       "       'hours_since_last_meal', 'id', 'image_type', 'is_thresholded',\n",
       "       'is_valid', 'map_type', 'mean_PDS_score', 'modality', 'modify_date',\n",
       "       'name', 'not_mni', 'number_of_subjects', 'perc_bad_voxels',\n",
       "       'perc_voxels_outside', 'race', 'reduced_representation',\n",
       "       'smoothness_fwhm', 'spsrq_score', 'statistic_parameters',\n",
       "       'subject_species', 'surface_left_file', 'surface_right_file',\n",
       "       'tanner_stage', 'target_template_image', 'thumbnail', 'url',\n",
       "       'waist_hip_ratio'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_images_df = get_images_metadata(collection_ids)\n",
    "all_images_df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The function returns a Pandas DataFrame with all metadata associated with the images. This way we can filter what images should go into our metaanalysis. We will only use single subject maps from this collection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BMI</th>\n",
       "      <th>age</th>\n",
       "      <th>brain_coverage</th>\n",
       "      <th>collection_id</th>\n",
       "      <th>file_size</th>\n",
       "      <th>id</th>\n",
       "      <th>number_of_subjects</th>\n",
       "      <th>perc_bad_voxels</th>\n",
       "      <th>perc_voxels_outside</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>67.00000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.0</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.0</td>\n",
       "      <td>122.000000</td>\n",
       "      <td>122.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>23.19194</td>\n",
       "      <td>33.516885</td>\n",
       "      <td>90.553975</td>\n",
       "      <td>3235.0</td>\n",
       "      <td>112867.745902</td>\n",
       "      <td>108309.500000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>76.678171</td>\n",
       "      <td>8.238074</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.93002</td>\n",
       "      <td>23.084441</td>\n",
       "      <td>2.953555</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4457.423544</td>\n",
       "      <td>35.362409</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.923337</td>\n",
       "      <td>2.965284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>19.52000</td>\n",
       "      <td>8.430000</td>\n",
       "      <td>80.609283</td>\n",
       "      <td>3235.0</td>\n",
       "      <td>102708.000000</td>\n",
       "      <td>108249.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>74.379708</td>\n",
       "      <td>2.867370</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>21.69500</td>\n",
       "      <td>14.737500</td>\n",
       "      <td>89.051625</td>\n",
       "      <td>3235.0</td>\n",
       "      <td>109903.000000</td>\n",
       "      <td>108279.250000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>76.130732</td>\n",
       "      <td>5.964508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>23.37000</td>\n",
       "      <td>26.615000</td>\n",
       "      <td>91.130076</td>\n",
       "      <td>3235.0</td>\n",
       "      <td>112553.500000</td>\n",
       "      <td>108309.500000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>76.698576</td>\n",
       "      <td>8.218514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>24.54000</td>\n",
       "      <td>59.022500</td>\n",
       "      <td>92.200343</td>\n",
       "      <td>3235.0</td>\n",
       "      <td>115561.750000</td>\n",
       "      <td>108339.750000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>77.239783</td>\n",
       "      <td>10.027860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>26.93000</td>\n",
       "      <td>76.060000</td>\n",
       "      <td>96.642300</td>\n",
       "      <td>3235.0</td>\n",
       "      <td>124164.000000</td>\n",
       "      <td>108370.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>78.883639</td>\n",
       "      <td>17.852171</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            BMI         age  brain_coverage  collection_id      file_size  \\\n",
       "count  67.00000  122.000000      122.000000          122.0     122.000000   \n",
       "mean   23.19194   33.516885       90.553975         3235.0  112867.745902   \n",
       "std     1.93002   23.084441        2.953555            0.0    4457.423544   \n",
       "min    19.52000    8.430000       80.609283         3235.0  102708.000000   \n",
       "25%    21.69500   14.737500       89.051625         3235.0  109903.000000   \n",
       "50%    23.37000   26.615000       91.130076         3235.0  112553.500000   \n",
       "75%    24.54000   59.022500       92.200343         3235.0  115561.750000   \n",
       "max    26.93000   76.060000       96.642300         3235.0  124164.000000   \n",
       "\n",
       "                  id  number_of_subjects  perc_bad_voxels  perc_voxels_outside  \n",
       "count     122.000000               122.0       122.000000           122.000000  \n",
       "mean   108309.500000                 1.0        76.678171             8.238074  \n",
       "std        35.362409                 0.0         0.923337             2.965284  \n",
       "min    108249.000000                 1.0        74.379708             2.867370  \n",
       "25%    108279.250000                 1.0        76.130732             5.964508  \n",
       "50%    108309.500000                 1.0        76.698576             8.218514  \n",
       "75%    108339.750000                 1.0        77.239783            10.027860  \n",
       "max    108370.000000                 1.0        78.883639            17.852171  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ss_images_df = all_images_df[all_images_df.analysis_level == 'single-subject']\n",
    "ss_images_df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performing the metaanalysis\n",
    "We begin by defining a helper function that takes a Dataframe of selected images and performs Stouffer's metaanalysis on them.\n",
    "\n",
    "**Disclaimer: the maps in this example are T maps which is technically incorrect for this technique. Without knowing the first level desing it is impossible to asses the degrees of freedom and convert them to Z maps, but for the sake of this excercise we will assume T == Z**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def perform_metaanalysis(images_df):\n",
    "    z_imgs = []\n",
    "    for i, row in images_df.iterrows():\n",
    "        download_url = row['file']\n",
    "        print(\"Downloading %s\"%download_url)\n",
    "        r = requests.get(download_url)\n",
    "        fp = io.BytesIO(r.content)\n",
    "        gzfileobj = GzipFile(filename=\"tmp.nii.gz\", mode='rb', fileobj=fp)\n",
    "        nii = nb.Nifti1Image.from_file_map({'image': nb.FileHolder(\"tmp.nii.gz\", gzfileobj)})\n",
    "\n",
    "        # making sure all images have the same size\n",
    "        resampled_nii = resample_to_img(nii, template_nii)\n",
    "        z_imgs.append(resampled_nii)\n",
    "\n",
    "    z_data = apply_mask(z_imgs, template_mask_nii)\n",
    "    results = stouffers(z_data, template_mask_nii, inference='ffx', null='theoretical', corr='FWE', two_sided=True)\n",
    "    return results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The above function downloads the selected maps and resamples them to match dimensions of the MNI template. Let's run it!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_TN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_TN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_TN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_TN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_TN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN009_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN052_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_EN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_EN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_EN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_TN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_TN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_TN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_TN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_TN009_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_TN011_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_TN012_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN009_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN011_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN013_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN014_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN016_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN018_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN019_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN021_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN023_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN024_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN025_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN026_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN027_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN009_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN010_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN011_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN012_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN013_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN014_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN016_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN017_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN020_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN021_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN022_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN027_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN029_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_CN030_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN010_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN011_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN012_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN013_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN014_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN015_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN016_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN017_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN018_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN019_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN020_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN021_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN022_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN023_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN006_H.nii.gz\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN009_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN010_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN011_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN012_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN013_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN014_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN015_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN016_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN017_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN018_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN019_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN020_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN021_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN025_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN026_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN027_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN028_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_TN030_H.nii.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\gdrive\\workspace\\megaanalysis\\src\\nimare\\nimare\\meta\\ibma\\ibma.py:240: RuntimeWarning: divide by zero encountered in log10\n"
     ]
    }
   ],
   "source": [
    "results = perform_metaanalysis(ss_images_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['z', 'p', 'log_p'])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.images.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "No we can plot the statistically significan ROIs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\me\\Miniconda3\\lib\\site-packages\\scipy\\ndimage\\measurements.py:272: DeprecationWarning: In future, it will be an error for 'np.bool_' scalars to be interpreted as an index\n",
      "  return _nd_image.find_objects(input, max_label)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<nilearn.plotting.displays.OrthoSlicer at 0x16348145be0>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 475.2x187.2 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_roi(math_img(formula = '(a < 0.05)*mask', a=results.images['p'], mask=template_mask_nii))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Filtering with metadata\n",
    "Because the DataFrame includes metadata we can use them to filter the images going into the metaanalysis. Lets showcase this, by redoing the above metaanalysis only with data from participants over 18 years old."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "ss_above_18_df = ss_images_df[ss_images_df.age > 18]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "68"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(ss_above_18_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_AN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_AB_EN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN009_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_AN052_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_EN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_EN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_EN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_GR_TN011_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN009_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN011_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN013_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN014_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN016_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN018_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN019_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN021_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN023_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN024_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN025_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN026_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_AN027_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN001_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN002_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN003_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN004_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN005_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN006_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN007_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN008_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN010_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN011_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN012_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN013_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN014_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN015_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN016_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN017_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN018_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN019_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN020_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN021_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN022_H.nii.gz\n",
      "Downloading https://neurovault.org/media/images/3235/spmT_0002_NL_EN023_H.nii.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\gdrive\\workspace\\megaanalysis\\src\\nimare\\nimare\\meta\\ibma\\ibma.py:240: RuntimeWarning: divide by zero encountered in log10\n",
      "C:\\Users\\me\\Miniconda3\\lib\\site-packages\\scipy\\ndimage\\measurements.py:272: DeprecationWarning: In future, it will be an error for 'np.bool_' scalars to be interpreted as an index\n",
      "  return _nd_image.find_objects(input, max_label)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<nilearn.plotting.displays.OrthoSlicer at 0x1630016d588>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 475.2x187.2 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "results = perform_metaanalysis(ss_above_18_df)\n",
    "plot_roi(math_img(formula = '(a < 0.05)*mask', a=results.images['p'], mask=template_mask_nii))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}