{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "\n# Brain segmentation with median_otsu\n\nWe show how to extract brain information and mask from a b0 image using DIPY_'s\n``segment.mask`` module.\n\nFirst import the necessary modules:\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "import numpy as np\nfrom dipy.data import get_fnames\nfrom dipy.io.image import load_nifti, save_nifti\nfrom dipy.segment.mask import median_otsu"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Download and read the data for this tutorial.\n\nThe ``scil_b0`` dataset contains different data from different companies and\nmodels. For this example, the data comes from a 1.5 Tesla Siemens MRI.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "data_fnames = get_fnames('scil_b0')\ndata, affine = load_nifti(data_fnames[1])\ndata = np.squeeze(data)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Segment the brain using DIPY's ``mask`` module.\n\n``median_otsu`` returns the segmented brain data and a binary mask of the\nbrain. It is possible to fine tune the parameters of ``median_otsu``\n(``median_radius`` and ``num_pass``) if extraction yields incorrect results\nbut the default parameters work well on most volumes. For this example,\nwe used 2 as ``median_radius`` and 1 as ``num_pass``\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "b0_mask, mask = median_otsu(data, median_radius=2, numpass=1)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Saving the segmentation results is very easy. We need the ``b0_mask``, and the\nbinary mask volumes. The affine matrix which transform the image's coordinates\nto the world coordinates is also needed. Here, we choose to save both images\nin ``float32``.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "fname = 'se_1.5t'\nsave_nifti(fname + '_binary_mask.nii.gz', mask.astype(np.float32), affine)\nsave_nifti(fname + '_mask.nii.gz', b0_mask.astype(np.float32), affine)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Quick view of the results middle slice using ``matplotlib``.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\nfrom dipy.core.histeq import histeq\n\nsli = data.shape[2] // 2\nplt.figure('Brain segmentation')\nplt.subplot(1, 2, 1).set_axis_off()\nplt.imshow(histeq(data[:, :, sli].astype('float')).T,\n           cmap='gray', origin='lower')\n\nplt.subplot(1, 2, 2).set_axis_off()\nplt.imshow(histeq(b0_mask[:, :, sli].astype('float')).T,\n           cmap='gray', origin='lower')\n\nplt.savefig('median_otsu.png')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        ".. figure:: median_otsu.png\n   :align: center\n\n   An application of median_otsu for brain segmentation.\n\n``median_otsu`` can also automatically crop the outputs to remove the largest\npossible number of background voxels. This makes outputted data significantly\nsmaller. Auto-cropping in ``median_otsu`` is activated by setting the\n``autocrop`` parameter to ``True``.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "b0_mask_crop, mask_crop = median_otsu(data, median_radius=4, numpass=4,\n                                      autocrop=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Saving cropped data using nibabel as demonstrated previously.\n\n.. include:: ../links_names.inc\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "save_nifti(fname + '_binary_mask_crop.nii.gz', mask_crop.astype(np.float32),\n           affine)\nsave_nifti(fname + '_mask_crop.nii.gz', b0_mask_crop.astype(np.float32),\n           affine)"
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.13"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}