slice-interpolation-for-zarpaint.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6bce021e",
   "metadata": {},
   "source": [
    "# Slice interpolation for labels in zarpaint"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "052bb9eb",
   "metadata": {},
   "source": [
    "## Resources\n",
    "\n",
    "* Paper:  Raya and Udupa \"Shape-Based Interpolation of Multidimensional Objects\" IEEE TRANSACTIONS ON MEDICAL IMAGING, VOL. 9, NO. I , MARCH 1990\n",
    "* [This StackOverflow post](https://stackoverflow.com/questions/48818373/interpolate-between-two-images) (not well executed for nd, but useful to read anyway)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d06c5e7",
   "metadata": {},
   "source": [
    "## Set up some example data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "0c66a7d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "6bdd9cbd",
   "metadata": {},
   "outputs": [],
   "source": [
    "import napari"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c88efc99",
   "metadata": {},
   "outputs": [],
   "source": [
    "viewer = napari.Viewer()\n",
    "# viewer.open_sample('napari', 'cells3d')\n",
    "# membrane_layer = viewer.layers[0]\n",
    "# nuclei_layer = viewer.layers[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "378903ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Image layer 'blobs' at 0x105ffddf0>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Traceback (most recent call last):\n",
      "  File \"/Users/genevieb/mambaforge/envs/zarpaint/lib/python3.9/site-packages/magicgui/widgets/_bases/value_widget.py\", line 57, in _on_value_change\n",
      "    self.changed.emit(value)\n",
      "  File \"psygnal/_signal.py\", line 682, in psygnal._signal.SignalInstance.emit\n",
      "  File \"psygnal/_signal.py\", line 724, in psygnal._signal.SignalInstance._run_emit_loop\n",
      "  File \"psygnal/_signal.py\", line 725, in psygnal._signal.SignalInstance._run_emit_loop\n",
      "  File \"psygnal/_signal.py\", line 745, in psygnal._signal.SignalInstance._run_emit_loop\n",
      "  File \"/Users/genevieb/mambaforge/envs/zarpaint/lib/python3.9/site-packages/toolz/functoolz.py\", line 306, in __call__\n",
      "    return self._partial(*args, **kwargs)\n",
      "  File \"/Users/genevieb/mambaforge/envs/zarpaint/lib/python3.9/site-packages/zarpaint/_zarpaint.py\", line 20, in _set_default_labels_path\n",
      "    source_image = source_image_event.value\n",
      "AttributeError: 'Image' object has no attribute 'value'\n"
     ]
    }
   ],
   "source": [
    "from skimage import data\n",
    "blobs = np.stack(\n",
    "    [\n",
    "        data.binary_blobs(\n",
    "            length=128, blob_size_fraction=0.05, n_dim=3, volume_fraction=f, seed=0\n",
    "        )\n",
    "        for f in np.linspace(0.05, 0.5, 10)\n",
    "    ],\n",
    "    axis=0,\n",
    ")\n",
    "viewer.add_image(blobs)\n",
    "\n",
    "# # numpy label data\n",
    "# single_nuclei_label = tifffile.imread('single-nuclei-label.tif')\n",
    "# viewer.add_labels(single_nuclei_label)\n",
    "# label_layer = viewer.layers[-1]\n",
    "# labels = label_layer.data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b3afda5",
   "metadata": {},
   "source": [
    "## Manually create labels with zarpaint GUI\n",
    "\n",
    "1. Plugins > zarpaint > create labels\n",
    "2. Create a new zarr file called \"mylabels.zarr\"\n",
    "3. Click \"Run\" in the zarpaint widget\n",
    "4. Label the same nucleus at two different slices (take note of which slice numbers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "a34808c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make note of which image slices you labelled\n",
    "slice_index_1 = 60\n",
    "slice_index_2 = 70\n",
    "interp_dim = 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1b279a2",
   "metadata": {},
   "source": [
    "## Exploring the tensorstore labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "119114bf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TensorStore({\n",
      "  'context': {\n",
      "    'cache_pool': {},\n",
      "    'data_copy_concurrency': {},\n",
      "    'file_io_concurrency': {},\n",
      "  },\n",
      "  'driver': 'zarr',\n",
      "  'dtype': 'uint32',\n",
      "  'kvstore': {'driver': 'file', 'path': 'blobls-test.zarr/'},\n",
      "  'metadata': {\n",
      "    'chunks': [1, 1, 128, 128],\n",
      "    'compressor': {\n",
      "      'blocksize': 0,\n",
      "      'clevel': 5,\n",
      "      'cname': 'lz4',\n",
      "      'id': 'blosc',\n",
      "      'shuffle': 1,\n",
      "    },\n",
      "    'dimension_separator': '.',\n",
      "    'dtype': '<u4',\n",
      "    'fill_value': 0,\n",
      "    'filters': None,\n",
      "    'order': 'C',\n",
      "    'shape': [10, 128, 128, 128],\n",
      "    'zarr_format': 2,\n",
      "  },\n",
      "  'transform': {\n",
      "    'input_exclusive_max': [[10], [128], [128], [128]],\n",
      "    'input_inclusive_min': [0, 0, 0, 0],\n",
      "  },\n",
      "})\n"
     ]
    }
   ],
   "source": [
    "# after you have done the manual steps above, run this cell\n",
    "label_layer = viewer.layers[-1]\n",
    "ts = label_layer.data  # tensorstore data for the labels\n",
    "print(ts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c8950929",
   "metadata": {},
   "outputs": [],
   "source": [
    "image_1 = np.take(label_layer.data, slice_index_1, axis=interp_dim)\n",
    "image_2 = np.take(label_layer.data, slice_index_2, axis=interp_dim)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a6c7ab8d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " ...\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]]\n",
      "\n",
      "numpy array values:\n",
      "\n",
      "[[[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " ...\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]\n",
      "\n",
      " [[0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  ...\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]\n",
      "  [0 0 0 ... 0 0 0]]]\n"
     ]
    }
   ],
   "source": [
    "print(image_1)\n",
    "print(\"\")\n",
    "print(\"numpy array values:\\n\")\n",
    "print(np.asarray(image_1))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "7fa99451",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2772\n",
      "3030\n"
     ]
    }
   ],
   "source": [
    "print(np.sum(image_1))\n",
    "print(np.sum(image_2))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "31b288bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x297a9cdc0>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "diff = image_1 - image_2\n",
    "\n",
    "fig, ax = plt.subplots(ncols=3, figsize=(10,3))\n",
    "ax[0].imshow(image_1[viewer.dims.current_step[0]])  # 4D original dataset\n",
    "ax[1].imshow(image_2[viewer.dims.current_step[0]])  # 4D original dataset\n",
    "ax[2].imshow(diff[viewer.dims.current_step[0]])     # 4D original dataset\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72200e8a",
   "metadata": {},
   "source": [
    "## Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "a3657b13",
   "metadata": {},
   "outputs": [],
   "source": [
    "# These functions are only used for this notebook, not for zarpaint\n",
    "# It's convenient to have an easy way to delete label scribbling\n",
    "\n",
    "def erase_slices(label_layer, slice_index_1, slice_index_2, interp_dim=0):\n",
    "    \"\"\"Erase label data on slices between slice_index_1 (inclusive),\n",
    "    and slice_index_2 (non-inclusive, same as the python slicing syntax).\"\"\"\n",
    "    for slice_number in range(slice_index_1, slice_index_2):\n",
    "        slices = [slice(None) for _ in label_layer.data.shape]\n",
    "        slices[interp_dim] = slice_number\n",
    "        label_layer.data[slices] = 0\n",
    "    label_layer.refresh()\n",
    "\n",
    "    \n",
    "def erase_all_slices(label_layer):\n",
    "    for i in label_layer.data.shape[0]:\n",
    "        label_layer.data[i, ...] = 0\n",
    "    label_layer.refresh()  # refresh current view"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ba7e762",
   "metadata": {},
   "source": [
    "These functions are for zarpaint..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "83658b31",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from scipy.interpolate import interpn\n",
    "from scipy.ndimage import distance_transform_edt\n",
    "\n",
    "\n",
    "def distance_transform(image):\n",
    "    \"\"\"Distance transform for a boolean image.\n",
    "    \n",
    "    Returns positive values inside the object,\n",
    "    and negative values outside.\n",
    "    \"\"\"\n",
    "    image = image.astype(bool)\n",
    "    edt = distance_transform_edt(image) - distance_transform_edt(~image)\n",
    "    return edt\n",
    "\n",
    "\n",
    "def point_and_values(image_1, image_2, interp_dim=0):\n",
    "    edt_1 = distance_transform(image_1)\n",
    "    edt_2 = distance_transform(image_2)\n",
    "    values = np.stack([edt_1, edt_2], axis=interp_dim)\n",
    "    points = tuple([np.arange(i) for i in values.shape])\n",
    "    return points, values\n",
    "\n",
    "\n",
    "def xi_coords(shape, percent=0.5):\n",
    "    slices = [slice(0, i) for i in shape]\n",
    "    xi = np.moveaxis(np.mgrid[slices], 0, -1).reshape(np.prod(shape), len(shape)).astype('float')\n",
    "    xi = np.insert(xi, interp_dim, percent, axis=1)\n",
    "    return xi\n",
    "\n",
    "\n",
    "def slice_iterator(slice_index_1, slice_index_2):\n",
    "    intermediate_slices = np.arange(slice_index_1 + 1, slice_index_2)\n",
    "    n_slices = slice_index_2 - slice_index_1 + 1  # inclusive\n",
    "    stepsize = 1 / n_slices\n",
    "    intermediate_percentages = np.arange(0 + stepsize, 1, stepsize)\n",
    "    return zip(intermediate_slices, intermediate_percentages)\n",
    "\n",
    "\n",
    "def interpolated_slice(percent, points, values, interp_dim=0, method='linear'):\n",
    "    # Find the original image shape\n",
    "    img_shape = list(values.shape)\n",
    "    del img_shape[interp_dim]\n",
    "    # Calculate the interpolated slice\n",
    "    xi = xi_coords(img_shape, percent=percent)\n",
    "    interpolated_img = interpn(points, values, xi, method=method)\n",
    "    interpolated_img = np.reshape(interpolated_img, img_shape) > 0\n",
    "    return interpolated_img\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3be6323c",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "\n",
    "\n",
    "# ## First draft, simply returns a numpy stack of the interpolated slices\n",
    "\n",
    "# def interpolate_between_slices(image_1, image_2, slice_index_1, slice_index_2, interp_dim=0, method='linear'):\n",
    "#     if slice_index_1 > slice_index_2:\n",
    "#         image_1, image_2 = image_2, image_1\n",
    "#         slice_index_1, slice_index_2 = slice_index_2, slice_index_1\n",
    "#     ####\n",
    "#     # possible extension, handle all label ids separately\n",
    "#     label_id = 1\n",
    "#     image_1 = image_1.astype(bool)\n",
    "#     image_2 = image_2.astype(bool)\n",
    "#     ####\n",
    "#     edt_1 = distance_transform(image_1)\n",
    "#     edt_2 = distance_transform(image_2)\n",
    "#     values = np.stack([edt_1, edt_2], axis=interp_dim)\n",
    "#     points = tuple([np.arange(i) for i in values.shape])\n",
    "\n",
    "#     output = []\n",
    "#     for slice_number, percentage in slice_iterator(slice_index_1, slice_index_2):\n",
    "#         interpolated_img = interpolated_slice(percentage, points, values, interp_dim=interp_dim, method='linear')\n",
    "#         output.append(interpolated_img)\n",
    "#     output = np.array(output)\n",
    "#     return output\n",
    "\n",
    "\n",
    "# result = interpolate_between_slices(image_1, image_2, slice_index_1, slice_index_2)\n",
    "# print(result.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "9b00f986",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Second draft, writes directly into tensorstore zarr array\n",
    "\n",
    "def interpolate_between_slices(label_layer, image_1, image_2, slice_index_1, slice_index_2, interp_dim=0, method='linear'):\n",
    "    if slice_index_1 > slice_index_2:\n",
    "        image_1, image_2 = image_2, image_1\n",
    "        slice_index_1, slice_index_2 = slice_index_2, slice_index_1\n",
    "    ####\n",
    "    # possible extension, handle all label ids separately\n",
    "    label_id = 1\n",
    "    image_1 = image_1.astype(bool)\n",
    "    image_2 = image_2.astype(bool)\n",
    "    ####\n",
    "    points, values = point_and_values(image_1, image_2, interp_dim=interp_dim)\n",
    "\n",
    "    for slice_number, percentage in slice_iterator(slice_index_1, slice_index_2):\n",
    "        interpolated_img = interpolated_slice(percentage, points, values, interp_dim=interp_dim, method='linear')\n",
    "        indices = [slice(None) for _ in range(label_layer.data.ndim)]\n",
    "        indices[interp_dim] = slice_number\n",
    "        indices = tuple(indices)\n",
    "        label_layer.data[indices][interpolated_img] = label_id\n",
    "    label_layer.refresh()  # will update the current view\n",
    "\n",
    "# interpolate_between_slices(label_layer, image_1, image_2, slice_index_1, slice_index_2)\n",
    "# print(\"Done!\")\n",
    "# print(\"Please scroll through napari to see the interpolated label slices\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b15cba0e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done!\n",
      "Please scroll through napari to see the interpolated label slices\n"
     ]
    }
   ],
   "source": [
    "interp_dim = 1\n",
    "interpolate_between_slices(label_layer, image_1, image_2, slice_index_1, slice_index_2, interp_dim=interp_dim)\n",
    "print(\"Done!\")\n",
    "print(\"Please scroll through napari to see the interpolated label slices\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8b3696b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "arr = np.random.random((512, 512, 512))\n",
    "arr.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6d737cd6",
   "metadata": {},
   "outputs": [],
   "source": [
    "label_layer.data[0, np.nonzero(interpolated_img)].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6653c0e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "viewer.dims.current_step"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bbf2cac8",
   "metadata": {},
   "outputs": [],
   "source": [
    "indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "224bfc53",
   "metadata": {},
   "outputs": [],
   "source": [
    "label_layer.data[5, 62, slice(None), slice(None)][:100, :100] = 2\n",
    "label_layer.refresh()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "509a792a",
   "metadata": {},
   "outputs": [],
   "source": [
    "label_layer.refresh()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9a1d05e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.full?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "56a44415",
   "metadata": {},
   "outputs": [],
   "source": [
    "label_layer.data[indices] = 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a803fbdc",
   "metadata": {},
   "outputs": [],
   "source": [
    "indices = list(np.nonzero(interpolated_img))\n",
    "indices.insert(interp_dim, np.full(len(indices[0]), slice_number))\n",
    "\n",
    "print(indices)\n",
    "print(len(indices))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dddd2ac4",
   "metadata": {},
   "outputs": [],
   "source": [
    "slices = tuple([slice(None) for _ in range(label_layer.data.ndim)])\n",
    "print(slices)\n",
    "label_layer.data.shape[slices]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6c8703f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorstore\n",
    "tensorstore.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bd047604",
   "metadata": {},
   "outputs": [],
   "source": [
    "!conda list | grep tensorstore"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "33e79bd3",
   "metadata": {},
   "outputs": [],
   "source": [
    "numpy.insert(interpolated_img, obj, values, axis=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b536a265",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.expand_dims?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a16e51dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "temp = np.expand_dims(interpolated_img, interp_dim)\n",
    "temp.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "05d2d439",
   "metadata": {},
   "outputs": [],
   "source": [
    "temp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0744b5a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "interp_dim = 1\n",
    "interpolate_between_slices(label_layer, image_1, image_2, slice_index_1, slice_index_2, interp_dim=interp_dim)\n",
    "print(\"Done!\")\n",
    "print(\"Please scroll through napari to see the interpolated label slices\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "af8aeb7b",
   "metadata": {},
   "outputs": [],
   "source": [
    "%debug"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "435f86d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "slices = [slice(None) for _ in range(label_layer.data.ndim)]\n",
    "print(slices)\n",
    "# label_layer.data[slices]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "60c159d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.take(label_layer.data, slice_number, axis=interp_dim).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bcaf0137",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "label_id = 1\n",
    "image_1 = image_1.astype(bool)\n",
    "image_2 = image_2.astype(bool)\n",
    "\n",
    "points, values = point_and_values(image_1, image_2, interp_dim=interp_dim)\n",
    "\n",
    "\n",
    "#for slice_number, percentage in slice_iterator(slice_index_1, slice_index_2):\n",
    "#    interpolated_img = interpolated_slice(percentage, points, values, interp_dim=interp_dim, method='linear')\n",
    "#    label_layer.data[slice_number, interpolated_img] = label_id\n",
    "#label_layer.refresh()  # will update the current view"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a911c2b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "percentage = 0.5\n",
    "percent = percentage\n",
    "\n",
    "img_shape = list(values.shape)\n",
    "del img_shape[interp_dim]\n",
    "# Calculate the interpolated slice\n",
    "\n",
    "# xi = xi_coords(img_shape, percent=percent)\n",
    "method='linear'\n",
    "shape = img_shape\n",
    "slices = [slice(0, i) for i in shape]\n",
    "xi = np.moveaxis(np.mgrid[slices], 0, -1).reshape(np.prod(shape), len(shape)).astype('float')\n",
    "xi = np.insert(xi, interp_dim, percent, axis=1)\n",
    "\n",
    "interpolated_img = interpn(points, values, xi, method=method)\n",
    "interpolated_img = np.reshape(interpolated_img, img_shape) > 0\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ee19afac",
   "metadata": {},
   "outputs": [],
   "source": [
    "interpolated_img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9857f388",
   "metadata": {},
   "outputs": [],
   "source": [
    "slice_number = 50\n",
    "temp = np.expand_dims(interpolated_img, interp_dim)\n",
    "print(temp.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c6c9ad3d",
   "metadata": {},
   "outputs": [],
   "source": [
    "temp\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b9f2a13e",
   "metadata": {},
   "outputs": [],
   "source": [
    "indices = np.argwhere(interpolated_img)\n",
    "indices = np.insert(indices, interp_dim, slice_number, axis=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "df3e8414",
   "metadata": {},
   "outputs": [],
   "source": [
    "xi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3d0c69ad",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ef63b9c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from scipy.interpolate import interpn\n",
    "from scipy.ndimage import distance_transform_edt\n",
    "\n",
    "\n",
    "def distance_transform(image):\n",
    "    \"\"\"Distance transform for a boolean image.\n",
    "    \n",
    "    Returns positive values inside the object,\n",
    "    and negative values outside.\n",
    "    \"\"\"\n",
    "    image = image.astype(bool)\n",
    "    edt = distance_transform_edt(image) - distance_transform_edt(~image)\n",
    "    return edt\n",
    "\n",
    "\n",
    "def point_and_values(image_1, image_2, interp_dim=0):\n",
    "    edt_1 = distance_transform(image_1)\n",
    "    edt_2 = distance_transform(image_2)\n",
    "    values = np.stack([edt_1, edt_2], axis=interp_dim)\n",
    "    points = tuple([np.arange(i) for i in values.shape])\n",
    "    return points, values\n",
    "\n",
    "\n",
    "def xi_coords(shape, percent=0.5):\n",
    "    slices = [slice(0, i) for i in shape]\n",
    "    xi = np.moveaxis(np.mgrid[slices], 0, -1).reshape(np.prod(shape), len(shape))\n",
    "    xi = xi = np.c_[np.full((np.prod(shape)), percent), xi]\n",
    "    return xi\n",
    "\n",
    "\n",
    "def slice_iterator(slice_index_1, slice_index_2):\n",
    "    intermediate_slices = np.arange(slice_index_1 + 1, slice_index_2)\n",
    "    n_slices = slice_index_2 - slice_index_1 + 1  # inclusive\n",
    "    stepsize = 1 / n_slices\n",
    "    intermediate_percentages = np.arange(0 + stepsize, 1, stepsize)\n",
    "    return zip(intermediate_slices, intermediate_percentages)\n",
    "\n",
    "\n",
    "def interpolated_slice(percent, points, values, interp_dim=0, method='linear'):\n",
    "    # Find the original image shape\n",
    "    img_shape = list(values.shape)\n",
    "    del img_shape[interp_dim]\n",
    "    # Calculate the interpolated slice\n",
    "    xi = xi_coords(img_shape, percent=percent)\n",
    "    interpolated_img = interpn(points, values, xi, method=method)\n",
    "    interpolated_img = np.reshape(interpolated_img, img_shape) > 0\n",
    "    return interpolated_img\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "33f94119",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f23e2b74",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d9c2522c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1e8d31e7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "449d7c17",
   "metadata": {},
   "source": [
    "## Sanity check\n",
    "\n",
    "Check that an interpolation percentage of zero returns a result equal to the input image_1.\n",
    "Check that an interpolation percentage of one returns a result equal to the input image_2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fc7528be",
   "metadata": {},
   "outputs": [],
   "source": [
    "interp_dim = 0\n",
    "label_id = 1\n",
    "\n",
    "points, values = point_and_values(image_1, image_2)\n",
    "\n",
    "output_1 = interpolated_slice(0.0, points, values, interp_dim=interp_dim)\n",
    "output_2 = interpolated_slice(1.0, points, values, interp_dim=interp_dim)\n",
    "\n",
    "assert np.allclose(image_1.astype(bool), interpolated_slice(0.0, points, values, interp_dim=interp_dim))\n",
    "assert np.allclose(image_2.astype(bool), interpolated_slice(1.0, points, values, interp_dim=interp_dim))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e7a9130",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "d9f0472d",
   "metadata": {},
   "source": [
    "## Test cases\n",
    "* a simple 3D test case (interpolating 2D slices)\n",
    "  * assert when percent = 0, the result exactly matches the input image_1\n",
    "  * assert when percent = 1.0, the result exactly matches the input image_2\n",
    "  * check results against a previous computation\n",
    "* A test case with multiple label id values\n",
    "* A 4D test case (interpolate across time points, from two labelled 3d blobs)\n",
    "* Check code is robust as to which order image_2 and image_2 are given (eg: user can be scrolling forwards OR backwards through the slices)\n",
    "* A test case when the structure is branching (one input blob turns into two output blobs)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "94872f6e",
   "metadata": {},
   "source": [
    "## TODO\n",
    "\n",
    "- [ ] Full docstrings for all functions\n",
    "- [ ] Basic test cases\n",
    "- [ ] Extend to multiple label id values\n",
    "- [ ] Consider extending functionality so we only need bring a small cropped section of each slice into memory, instead of needing to convert the full slice to numpy\n",
    "- [ ] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5ab43dfe",
   "metadata": {},
   "outputs": [],
   "source": [
    "# answers\n",
    "viewer.dims.not_displayed  # how to get the dimension that is a slider bar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f5c9a76d",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c8ab9c70",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fac9732a",
   "metadata": {},
   "outputs": [],
   "source": [
    "viewer.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "91c24ec4",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "e8f0b44cf16a59ebfa2ed0b00c97bf93787b28c44f2aafb01610def517855b78"
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}