Direct_Optimization_20211211.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Direct Optimization 20211211.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "V5yTbigaU1EI",
        "kThxybhdVAnq",
        "yb2m-Z0cDgeq",
        "gtubIUZqU3pJ",
        "iND6vl0BU6fj",
        "w68n5geH84F8",
        "wHBBFxELPKV2"
      ],
      "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/surajsrivathsa/image_registration/blob/main/Direct_Optimization_20211211.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V5yTbigaU1EI"
      },
      "source": [
        "# Preamble"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "M6X8Kc7sTu30",
        "outputId": "cfe7aac5-90a8-47f2-8266-36ab856dea8b"
      },
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.utils.data as Data\n",
        "import torch.optim as optim\n",
        "import torchvision\n",
        "from torchvision import datasets, transforms\n",
        "from matplotlib import pyplot as plt\n",
        "from scipy.ndimage.interpolation import rotate\n",
        "import nibabel as nb\n",
        "from google.colab import drive\n",
        "import glob\n",
        "import os, sys\n",
        "drive.mount('/content/drive',  force_remount=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kThxybhdVAnq"
      },
      "source": [
        "# Data Reading"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gfyBUS1-CFLK"
      },
      "source": [
        "def load_3D( name):\n",
        "    # model_np = np.zeros(shape=(160, 192, 224))\n",
        "    # X_nb = nb.load(name)\n",
        "    # X_np = X_nb.dataobj\n",
        "    # #print(\"Oreintation: {}\".format(nb.aff2axcodes(X_nb.affine)))\n",
        "    # model_np[:, :, :] = X_np[42:202, 32:224, 16:240]\n",
        "    # #model_np = np.reshape(model_np, (1,)+ model_np.shape)\n",
        "    # return model_np\n",
        "    resamplng_shape = (128, 128, 128)\n",
        "    \n",
        "\n",
        "    X_nb = nb.load(name)\n",
        "    X_np = X_nb.dataobj\n",
        "    least_intensity = np.min(X_np)\n",
        "    model_np = np.full(shape=resamplng_shape, fill_value=least_intensity)\n",
        "\n",
        "    #X_np = imgnorm(X_np_old)\n",
        "    x_dim, y_dim, z_dim = X_np.shape\n",
        "    x_ltail = (resamplng_shape[0] - x_dim)//2 \n",
        "    y_ltail = (resamplng_shape[1] - y_dim)//2\n",
        "    z_ltail = (resamplng_shape[2] - z_dim)//2\n",
        "\n",
        "    x_rtail = resamplng_shape[0] - x_ltail - 1\n",
        "    y_rtail = resamplng_shape[1] - y_ltail - 1\n",
        "    z_rtail = resamplng_shape[2] - z_ltail - 1\n",
        "    #print(\"Oreintation: {}\".format(nb.aff2axcodes(X_nb.affine)))\n",
        "    #model_np[:, :, :] = X_np[42:202, 32:224, 16:240]\n",
        "    model_np[x_ltail:x_rtail, y_ltail:y_rtail, z_ltail:z_rtail] = X_np[:, :, :]\n",
        "    model_np = np.reshape(model_np, (1,)+ model_np.shape)\n",
        "    model_np = np.reshape(model_np, (1,)+ model_np.shape)\n",
        "    #myimg = imgnorm(model_np)\n",
        "    return model_np\n",
        "\n",
        "def load_4D(name):\n",
        "    # (256, 256, 256)\n",
        "    model_np = np.zeros(shape=(128, 128, 128))\n",
        "    X_nb = nb.load(name)\n",
        "    \n",
        "    X_np = X_nb.dataobj\n",
        "    #print(\"Oreintation: {}\".format(nb.aff2axcodes(X_nb.affine)))\n",
        "    #model_np[:, :, 0:X_np.shape[2]] = X_np[0:128, 0:128, :]\n",
        "    #model_np = np.reshape(model_np, (1,)+ model_np.shape)\n",
        "    model_np = np.reshape(X_np, (1,)+ X_np.shape)\n",
        "    model_np = np.reshape(model_np, (1,)+ model_np.shape)\n",
        "    return model_np\n",
        "\n",
        "def imgnorm(N_I,index1=0.0001,index2=0.0001):\n",
        "    I_sort = np.sort(N_I.flatten())\n",
        "    I_min = I_sort[int(index1*len(I_sort))]\n",
        "    I_max = I_sort[-int(index2*len(I_sort))]\n",
        "    N_I =1.0*(N_I-I_min)/(I_max-I_min)\n",
        "    N_I[N_I>1.0]=1.0\n",
        "    N_I[N_I<0.0]=0.0\n",
        "    N_I2 = N_I.astype(np.float32)\n",
        "    return N_I2\n",
        "\n",
        "\n",
        "def Norm_Zscore( img):\n",
        "    img= (img-np.mean(img))/np.std(img) \n",
        "    return img\n",
        "\n",
        "\n",
        "def run3functions( fp):\n",
        "    myimg = load_4D(fp)\n",
        "    myimg2 = imgnorm(myimg)\n",
        "    return myimg2\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BDg3BDUYkSaJ"
      },
      "source": [
        "# data_path_fixed = \"/content/drive/MyDrive/DATASET_FINAL/Test/intramodal/T1/pairwise_registered/Batch_1\"\n",
        "# data_path_moving = \"/content/drive/MyDrive/DATASET_FINAL/Test/intramodal/T1/pairwise_registered/Batch_1_F_Batch_2_M\"\n",
        "# data_path_warped = '/content/drive/MyDrive/DATASET_FINAL/Test/intramodal/T1/pairwise_registered/direct_optimization_adamw_registered_Batch1_F_Batch_2_M'\n",
        "data_path_fixed = \"/content/drive/MyDrive/DATASET_FINAL/Test/intermodal/Batch_1/T1_F_T2_M/T1\"\n",
        "data_path_moving = \"/content/drive/MyDrive/DATASET_FINAL/Test/intermodal/Batch_1/T1_F_T2_M/T2_affine\"\n",
        "data_path_warped = '/content/drive/MyDrive/DATASET_FINAL/Test/intermodal/Batch_1/direct_optimization_sgd_T1_F_T2_M'\n",
        "data_path_lst_fixed = sorted(glob.glob(os.path.join(data_path_fixed, \"*.nii.gz\")))\n",
        "data_path_lst_fixed = data_path_lst_fixed[:3]\n",
        "data_path_lst_moving = sorted(glob.glob(os.path.join(data_path_moving, \"*.nii.gz\")))\n",
        "data_path_lst_moving = data_path_lst_moving[:3]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "j25tUDXVmJ1a"
      },
      "source": [
        "- Create a list of deformation fields with same number as fixed images\n",
        "- Read fixed images and put the tensors into a list\n",
        "- Read moving images and put the tensors into a list\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M2ynzljYljLh"
      },
      "source": [
        "deformation_field_tnsr_lst = [ ]\n",
        "fixed_img_tnsr_lst = []\n",
        "moving_img_tnsr_lst = []\n",
        "for i in range(len(data_path_lst_fixed)):\n",
        "  deformation_field_tnsr = torch.randn(size = (1,3,128,128,128)).to(\"cuda\") * 0.00001\n",
        "  deformation_field_tnsr.requires_grad = True\n",
        "  deformation_field_tnsr_lst.append(deformation_field_tnsr)\n",
        "\n",
        "  fixed_img_tnsr = torch.from_numpy(run3functions(data_path_lst_fixed[i])).to(\"cuda\")\n",
        "  fixed_img_tnsr.requires_grad = False\n",
        "  fixed_img_tnsr_lst.append(fixed_img_tnsr)\n",
        "\n",
        "  moving_img_tnsr = torch.from_numpy(run3functions(data_path_lst_moving[i])).to(\"cuda\")\n",
        "  moving_img_tnsr.requires_grad = False\n",
        "  moving_img_tnsr_lst.append(moving_img_tnsr)\n",
        "\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yb2m-Z0cDgeq"
      },
      "source": [
        "# Deformation field integration"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r9fiuu9MDhMd"
      },
      "source": [
        "class VecInt(nn.Module):\n",
        "    \"\"\"\n",
        "    Integrates a vector field via scaling and squaring.\n",
        "    \"\"\"\n",
        "\n",
        "    def __init__(self, inshape, nsteps):\n",
        "        super().__init__()\n",
        "        \n",
        "        assert nsteps >= 0, 'nsteps should be >= 0, found: %d' % nsteps\n",
        "        self.nsteps = nsteps\n",
        "        self.scale = 1.0 / (2 ** self.nsteps)\n",
        "        self.transformer = SpatialTransformer(size = inshape)\n",
        "\n",
        "    def forward(self, vec):\n",
        "        vec = vec * self.scale\n",
        "        for _ in range(self.nsteps):\n",
        "            vec = vec + self.transformer(vec, vec)\n",
        "        return vec\n",
        "\n",
        "\n",
        "class ResizeTransform(nn.Module):\n",
        "    \"\"\"\n",
        "    Resize a transform, which involves resizing the vector field *and* rescaling it.\n",
        "    \"\"\"\n",
        "\n",
        "    def __init__(self, vel_resize, ndims):\n",
        "        super().__init__()\n",
        "        self.factor = 1.0 / vel_resize\n",
        "        self.mode = 'linear'\n",
        "        if ndims == 2:\n",
        "            self.mode = 'bi' + self.mode\n",
        "        elif ndims == 3:\n",
        "            self.mode = 'tri' + self.mode\n",
        "\n",
        "    def forward(self, x):\n",
        "        if self.factor < 1:\n",
        "            # resize first to save memory\n",
        "            x = F.interpolate(x, align_corners=True, scale_factor=self.factor, mode=self.mode, recompute_scale_factor=True)\n",
        "            x = self.factor * x\n",
        "\n",
        "        elif self.factor > 1:\n",
        "            # multiply first to save memory\n",
        "            x = self.factor * x\n",
        "            x = F.interpolate(x, align_corners=True, scale_factor=self.factor, mode=self.mode, recompute_scale_factor=True)\n",
        "\n",
        "        # don't do anything if resize is 1\n",
        "        return x"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gtubIUZqU3pJ"
      },
      "source": [
        "# Losses"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M6VWjCo3U6AT"
      },
      "source": [
        "class NormalizedCrossCorrelation(nn.Module):\n",
        "    \"\"\" N-dimensional normalized cross correlation (NCC)\n",
        "    Args:\n",
        "        eps (float, optional): Epsilon value for numerical stability. Defaults to 1e-8.\n",
        "        return_map (bool, optional): If True, also return the correlation map. Defaults to False.\n",
        "        reduction (str, optional): Specifies the reduction to apply to the output:\n",
        "            ``'mean'`` | ``'sum'``. Defaults to ``'mean'``.\n",
        "    \"\"\"\n",
        "    def __init__(self,\n",
        "                 eps=1e-8,\n",
        "                 return_map=False,\n",
        "                 reduction='mean'):\n",
        "\n",
        "        super(NormalizedCrossCorrelation, self).__init__()\n",
        "\n",
        "        self._eps = eps\n",
        "        self._return_map = return_map\n",
        "        self._reduction = reduction\n",
        "\n",
        "    def forward(self, x, y):\n",
        "\n",
        "        return normalized_cross_correlation(x, y,self._return_map, self._reduction, self._eps)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MZz6-oQ7U6D-"
      },
      "source": [
        "def normalized_cross_correlation(x, y, return_map, reduction='mean', eps=1e-8):\n",
        "    \"\"\" N-dimensional normalized cross correlation (NCC)\n",
        "    Args:\n",
        "        x (~torch.Tensor): Input tensor.\n",
        "        y (~torch.Tensor): Input tensor.\n",
        "        return_map (bool): If True, also return the correlation map.\n",
        "        reduction (str, optional): Specifies the reduction to apply to the output:\n",
        "            ``'mean'`` | ``'sum'``. Defaults to ``'sum'``.\n",
        "        eps (float, optional): Epsilon value for numerical stability. Defaults to 1e-8.\n",
        "    Returns:\n",
        "        ~torch.Tensor: Output scalar\n",
        "        ~torch.Tensor: Output tensor\n",
        "    \"\"\"\n",
        "\n",
        "    shape = x.shape\n",
        "    b = shape[0]\n",
        "\n",
        "    # reshape\n",
        "    x = x.view(b, -1)\n",
        "    y = y.view(b, -1)\n",
        "\n",
        "    # mean\n",
        "    x_mean = torch.mean(x, dim=1, keepdim=True)\n",
        "    y_mean = torch.mean(y, dim=1, keepdim=True)\n",
        "\n",
        "    # deviation\n",
        "    x = x - x_mean\n",
        "    y = y - y_mean\n",
        "\n",
        "    dev_xy = torch.mul(x,y)\n",
        "    dev_xx = torch.mul(x,x)\n",
        "    dev_yy = torch.mul(y,y)\n",
        "\n",
        "    dev_xx_sum = torch.sum(dev_xx, dim=1, keepdim=True)\n",
        "    dev_yy_sum = torch.sum(dev_yy, dim=1, keepdim=True)\n",
        "\n",
        "    ncc = torch.div(dev_xy + eps / dev_xy.shape[1],\n",
        "                    torch.sqrt( torch.mul(dev_xx_sum, dev_yy_sum)) + eps)\n",
        "    ncc_map = ncc.view(b, *shape[1:])\n",
        "\n",
        "    # reduce\n",
        "    if reduction == 'mean':\n",
        "        ncc = torch.mean(torch.sum(ncc, dim=1))\n",
        "    elif reduction == 'sum':\n",
        "        ncc = torch.sum(ncc)\n",
        "    else:\n",
        "        raise KeyError('unsupported reduction type: %s' % reduction)\n",
        "\n",
        "    if not return_map:\n",
        "        return ncc\n",
        "    \n",
        "    if (torch.isclose(torch.tensor([-1.0]).to(\"cuda\"), ncc).any()):\n",
        "      ncc = ncc + torch.tensor([0.01]).to(\"cuda\")\n",
        "\n",
        "    elif (torch.isclose(torch.tensor([1.0]).to(\"cuda\"), ncc).any()):\n",
        "      ncc = ncc - torch.tensor([0.01]).to(\"cuda\")\n",
        "\n",
        "    return ncc, ncc_map\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LBbo3fP5WVem"
      },
      "source": [
        "def mae_loss(input, target):\n",
        "    y_true_f = input.view(-1)\n",
        "    y_pred_f = target.view(-1)\n",
        "    diff = torch.abs(y_true_f-y_pred_f)\n",
        "    mae = diff.mean() \n",
        "    return mae"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kql7udyrWVhg"
      },
      "source": [
        "class Grad:\n",
        "    \"\"\"\n",
        "    N-D gradient loss.\n",
        "    \"\"\"\n",
        "\n",
        "    def __init__(self, penalty='l1', loss_mult=None):\n",
        "        self.penalty = penalty\n",
        "        self.loss_mult = loss_mult\n",
        "        super(Grad, self).__init__()\n",
        "\n",
        "    def loss(self, _, y_pred):\n",
        "        dy = torch.abs(y_pred[:, :, 1:, :, :] - y_pred[:, :, :-1, :, :])\n",
        "        dx = torch.abs(y_pred[:, :, :, 1:, :] - y_pred[:, :, :, :-1, :])\n",
        "        dz = torch.abs(y_pred[:, :, :, :, 1:] - y_pred[:, :, :, :, :-1])\n",
        "\n",
        "        if self.penalty == 'l2':\n",
        "            dy = dy * dy\n",
        "            dx = dx * dx\n",
        "            dz = dz * dz\n",
        "\n",
        "        d = torch.mean(dx) + torch.mean(dy) + torch.mean(dz)\n",
        "        grad = d / 3.0\n",
        "\n",
        "        if self.loss_mult is not None:\n",
        "            grad *= self.loss_mult\n",
        "        return grad"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oJRZEadTFay7"
      },
      "source": [
        "\n",
        "\n",
        "class MutualInformation(nn.Module):\n",
        "\n",
        "\tdef __init__(self, sigma=0.4, num_bins=30, normalize=True):\n",
        "\t\tsuper(MutualInformation, self).__init__()\n",
        "\n",
        "\t\tself.sigma = 2*sigma**2\n",
        "\t\tself.num_bins = num_bins\n",
        "\t\tself.normalize = normalize\n",
        "\t\tself.epsilon = 1e-10\n",
        "\n",
        "\t\tself.bins = nn.Parameter(torch.linspace(0, 255.0, num_bins, device='cuda').float(), requires_grad=True)\n",
        "\n",
        "\n",
        "\tdef marginalPdf(self, values):\n",
        "\n",
        "\t\tresiduals = values - self.bins.unsqueeze(0).unsqueeze(0)\n",
        "\t\tkernel_values = torch.exp(-0.5*(residuals / self.sigma).pow(2))\n",
        "\t\t\n",
        "\t\tpdf = torch.mean(kernel_values, dim=1)\n",
        "\t\tnormalization = torch.sum(pdf, dim=1).unsqueeze(1) + self.epsilon\n",
        "\t\tpdf = pdf / normalization\n",
        "\t\t\n",
        "\t\treturn pdf, kernel_values\n",
        "\n",
        "\n",
        "\tdef jointPdf(self, kernel_values1, kernel_values2):\n",
        "\n",
        "\t\tjoint_kernel_values = torch.matmul(kernel_values1.transpose(1, 2), kernel_values2) \n",
        "\t\tnormalization = torch.sum(joint_kernel_values, dim=(1,2)).view(-1, 1, 1) + self.epsilon\n",
        "\t\tpdf = joint_kernel_values / normalization\n",
        "\n",
        "\t\treturn pdf\n",
        "\n",
        "\n",
        "\tdef getMutualInformation(self, input1, input2):\n",
        "\t\t'''\n",
        "\t\t\tinput1: B, C, H, W, D\n",
        "\t\t\tinput2: B, C, H, W, D\n",
        "\t\t\treturn: scalar\n",
        "\t\t'''\n",
        "\n",
        "\t\t# Torch tensors for images between (0, 1)\n",
        "\t\tinput1 = input1*255\n",
        "\t\tinput2 = input2*255\n",
        "\n",
        "\t\tB, C, H, W, D = input1.shape\n",
        "\n",
        "\t\tx1 = input1.view(B, H*W*D, C)\n",
        "\t\tx2 = input2.view(B, H*W*D, C)\n",
        "\t\t\n",
        "\t\tpdf_x1, kernel_values1 = self.marginalPdf(x1)\n",
        "\t\tpdf_x2, kernel_values2 = self.marginalPdf(x2)\n",
        "\t\tpdf_x1x2 = self.jointPdf(kernel_values1, kernel_values2)\n",
        "\n",
        "\t\tH_x1 = -torch.sum(pdf_x1*torch.log2(pdf_x1 + self.epsilon), dim=1)\n",
        "\t\tH_x2 = -torch.sum(pdf_x2*torch.log2(pdf_x2 + self.epsilon), dim=1)\n",
        "\t\tH_x1x2 = -torch.sum(pdf_x1x2*torch.log2(pdf_x1x2 + self.epsilon), dim=(1,2))\n",
        "\n",
        "\t\tmutual_information = H_x1 + H_x2 - H_x1x2\n",
        "\t\t\n",
        "\t\tif self.normalize:\n",
        "\t\t\tmutual_information = 2*mutual_information/(H_x1+H_x2)\n",
        "\n",
        "\t\treturn mutual_information\n",
        "\n",
        "\n",
        "\tdef forward(self, input1, input2):\n",
        "\t\t'''\n",
        "\t\t\tinput1: B, C, H, W\n",
        "\t\t\tinput2: B, C, H, W\n",
        "\t\t\treturn: scalar\n",
        "\t\t'''\n",
        "\t\treturn self.getMutualInformation(input1, input2)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cHna-N18HSR5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ab48d220-1642-4c9b-e076-d9e6016b56c3"
      },
      "source": [
        "nmi = MutualInformation()\n",
        "nmi_similarity = nmi(fixed_img_tnsr_lst[0].contiguous(),moving_img_tnsr_lst[2].contiguous() )\n",
        "nmi_similarity"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([1.5519], device='cuda:0', grad_fn=<DivBackward0>)"
            ]
          },
          "metadata": {},
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iND6vl0BU6fj"
      },
      "source": [
        "# Spatial Transformer"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1R863r0qVLUi"
      },
      "source": [
        "class SpatialTransformer(nn.Module):\n",
        "    \"\"\"\n",
        "    N-D Spatial Transformer\n",
        "    \"\"\"\n",
        "\n",
        "    def __init__(self, size, is_affine=False, theta = None, mode='bilinear', affine_image_size =  (1, 1, 128, 128, 128)):\n",
        "        super().__init__()\n",
        "\n",
        "        self.mode = mode\n",
        "        self.isaffine = is_affine\n",
        "        self.theta = theta\n",
        "        self.affine_image_size =  affine_image_size\n",
        "        # create sampling grid\n",
        "        # registering the grid as a buffer cleanly moves it to the GPU, but it also\n",
        "        # adds it to the state dict. this is annoying since everything in the state dict\n",
        "        # is included when saving weights to disk, so the model files are way bigger\n",
        "        # than they need to be. so far, there does not appear to be an elegant solution.\n",
        "        # see: https://discuss.pytorch.org/t/how-to-register-buffer-without-polluting-state-dict\n",
        "\n",
        "        if (self.isaffine):\n",
        "          grid = F.affine_grid(self.theta, self.affine_image_size, align_corners=False)\n",
        "          #grid = grid.permute(0, 4, 1, 2, 3)\n",
        "          self.register_buffer('grid', grid)\n",
        "        else:\n",
        "          vectors = [torch.arange(0, s) for s in size]\n",
        "          grids = torch.meshgrid(vectors)\n",
        "          grid = torch.stack(grids)\n",
        "          grid = torch.unsqueeze(grid, 0)\n",
        "          grid = grid.type(torch.FloatTensor)\n",
        "          self.register_buffer('grid', grid)\n",
        "          # print(grid)\n",
        "\n",
        "    def forward(self, src, flow=None):      \n",
        "      if (self.isaffine):\n",
        "        grid = F.affine_grid(self.theta, self.affine_image_size)    \n",
        "        # print(grid)    \n",
        "        warped_image = F.grid_sample(src, grid)\n",
        "        #warped_image = warped_image.permute(0, 4, 1, 2, 3)\n",
        "        return warped_image\n",
        "      else:\n",
        "        # new locations\n",
        "        # print(grid)\n",
        "        new_locs = self.grid + flow\n",
        "        shape = flow.shape[2:]\n",
        "\n",
        "        # need to normalize grid values to [-1, 1] for resampler\n",
        "        for i in range(len(shape)):\n",
        "            new_locs[:, i, ...] = 2 * (new_locs[:, i, ...] / (shape[i] - 1) - 0.5)\n",
        "\n",
        "        # move channels dim to last position\n",
        "        # also not sure why, but the channels need to be reversed\n",
        "        if len(shape) == 2:\n",
        "            new_locs = new_locs.permute(0, 2, 3, 1)\n",
        "            new_locs = new_locs[..., [1, 0]]\n",
        "        elif len(shape) == 3:\n",
        "            new_locs = new_locs.permute(0, 2, 3, 4, 1)\n",
        "            new_locs = new_locs[..., [2, 1, 0]]\n",
        "\n",
        "        return F.grid_sample(src, new_locs, align_corners=True, mode=self.mode)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h3JjPkJqVLaH"
      },
      "source": [
        "spatial_transformer_deformable = SpatialTransformer(size=(128, 128, 128), is_affine=False, affine_image_size =  (1, 1, 128, 128, 128)).to(\"cuda\")\n",
        "for param in spatial_transformer_deformable.parameters():\n",
        "  param.requires_grad = False\n",
        "  param.volatile=True"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s8a6YmLkVchN"
      },
      "source": [
        "# Training"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qXgU9Wh7oBP6"
      },
      "source": [
        "initialize losses"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "guuteOOMaREn"
      },
      "source": [
        "smoothness_loss = Grad(penalty='l2')\n",
        "# similarity_loss = NormalizedCrossCorrelation().to(\"cuda\")\n",
        "similarity_loss = MutualInformation().to(\"cuda\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wLsfcBg5oItk"
      },
      "source": [
        "initialize optimizers and learning rate"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VgvYk-21qRrD"
      },
      "source": [
        "adadelta_learning_rate = 1e-3\n",
        "adagrad_learning_rate = 1e-3\n",
        "adam_learning_rate = 1e-3\n",
        "adamw_learning_rate = 1e-3\n",
        "rmsprop_learning_rate = 1e-3\n",
        "sgd_learning_rate = 1e-3"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l7enMSXTqAc_"
      },
      "source": [
        "optimizer_dict = {'optimizer_adadelta': [ torch.optim.Adadelta([deformation_field], lr=adadelta_learning_rate) for deformation_field in deformation_field_tnsr_lst],\n",
        "                  'optimizer_adagrad': [ torch.optim.Adagrad([deformation_field], lr=adagrad_learning_rate) for deformation_field in deformation_field_tnsr_lst],\n",
        "                  'optimizer_adam': [ torch.optim.Adam([deformation_field], lr=adam_learning_rate) for deformation_field in deformation_field_tnsr_lst],\n",
        "                  'optimizer_adamw': [ torch.optim.Adam([deformation_field], lr=adamw_learning_rate) for deformation_field in deformation_field_tnsr_lst],\n",
        "                  'optimizer_rmsprop': [ torch.optim.RMSprop([deformation_field], lr=rmsprop_learning_rate) for deformation_field in deformation_field_tnsr_lst],\n",
        "                  'optimizer_sgd': [ torch.optim.SGD([deformation_field], lr=sgd_learning_rate) for deformation_field in deformation_field_tnsr_lst]\n",
        "                  }"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tloIQTgTunhE"
      },
      "source": [
        "scheduler_lst = [torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma=0.99) for optimizer in optimizer_dict['optimizer_adamw']]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8uE8PBKptqeQ"
      },
      "source": [
        "Adding vector integration code, this enforces diffeomorphic transform"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SxWpUFhCtq3Q"
      },
      "source": [
        "dim = 3\n",
        "int_downsize = 2\n",
        "down_shape = [int(dim/ int_downsize) for dim in (128, 128, 128)]\n",
        "resize = ResizeTransform(2, 3).to(\"cuda\")\n",
        "fullsize = ResizeTransform(0.5, 3).to(\"cuda\")\n",
        "integrate = VecInt(down_shape, 7).to(\"cuda\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "92N5wUsZ9GGY"
      },
      "source": [
        "function that optimizers for one image pair"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QNkptfFRxSEX"
      },
      "source": [
        "def optimize_one_image_pair(fixed_img_tnsr, moving_img_tnsr, deformation_field, optimizer = optimizer_dict['optimizer_adam'][0], scheduler = scheduler_lst[0], epochs=1001, decay_flag=False):\n",
        "  optimized_deformation_field_lst = []\n",
        "  loss_lst = []\n",
        "  for step in range(epochs):\n",
        "    \n",
        "    # Find warped image given moving image from spatial transformer\n",
        "    pos_flow = resize(deformation_field)\n",
        "    integrated_pos_flow = integrate(pos_flow)\n",
        "    full_flow = fullsize(integrated_pos_flow)\n",
        "    warped_image_tensor = spatial_transformer_deformable(moving_img_tnsr.contiguous(), full_flow)\n",
        "    \n",
        "    # Find loss between warped image and fixed image\n",
        "    similarity_loss_ncc = 1.0 * similarity_loss(fixed_img_tnsr.contiguous(), warped_image_tensor.contiguous())\n",
        "    sm_loss =  0.5 * smoothness_loss.loss(\"\",deformation_field)\n",
        "    #  similarity_loss_mae = 0.1 * mae_loss(fixed_img_tnsr.contiguous(), warped_image_tensor.contiguous())\n",
        "    similarity_loss_mae = 0.0\n",
        "    total_loss = similarity_loss_ncc + sm_loss # + similarity_loss_mae\n",
        "\n",
        "    # Backpropagate loss through network\n",
        "    optimizer.zero_grad()          \n",
        "    total_loss.backward()\n",
        "    optimizer.step()\n",
        "\n",
        "    # Decay if the falg is true\n",
        "    if decay_flag:\n",
        "      scheduler.step()\n",
        "\n",
        "    # Log losses\n",
        "    if (step % 500 == 0):\n",
        "      # loss_lst.append(step, similarity_loss_ncc, sm_loss, similarity_loss_mae, total_loss)\n",
        "      print(\"losses at {} are {}, {}, {} and {}\".format(step, similarity_loss_ncc, sm_loss, similarity_loss_mae, total_loss))\n",
        "\n",
        "    # Save deformation field every multiple of 500 epoch that is 500,1000,1500,2000\n",
        "    if (step % 500 == 0) and (step > 0):\n",
        "      # deformation_field_detached = deformation_field.detach()\n",
        "      # optimized_deformation_field_lst.append(deformation_field_detached)\n",
        "      pass\n",
        "    \n",
        "    # delete unused warped image to save memeory\n",
        "    del warped_image_tensor\n",
        "\n",
        "  # Save deformation field\n",
        "  deformation_field_detached = deformation_field.detach()\n",
        "  optimized_deformation_field_lst.append(deformation_field_detached)\n",
        "\n",
        "  # clear memory of deformation field\n",
        "  del deformation_field\n",
        "\n",
        "  return (optimized_deformation_field_lst, loss_lst)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P7hriLRl9AZR"
      },
      "source": [
        "function that saves the warped images for the list of deformation fields in the folder path"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bIfUI7HD4Xyv"
      },
      "source": [
        "def generate_and_save_warped_images( fixed_img_path, moving_img_path, fixed_img_tnsr, moving_img_tnsr, optimized_deformation_field_lst, data_path_warped=data_path_warped):\n",
        "  fixed_image_name = os.path.basename(fixed_img_path).replace('.nii.gz', '')\n",
        "  moving_image_name = os.path.basename(moving_img_path).replace('.nii.gz', '')\n",
        "  warped_image_name = fixed_image_name + \"|\" + moving_image_name\n",
        "  counter = 1\n",
        "  for deformation_field in optimized_deformation_field_lst:\n",
        "    pos_flow = resize(deformation_field)\n",
        "    integrated_pos_flow = integrate(pos_flow)\n",
        "    full_flow = fullsize(integrated_pos_flow)\n",
        "    warped_img_tnsr = spatial_transformer_deformable(moving_img_tnsr, full_flow)\n",
        "    warped_img_np = warped_img_tnsr.to(\"cpu\").numpy()\n",
        "    warped_img_nb = nb.Nifti1Image(warped_img_np[0,0,:,:,:], np.eye(4))\n",
        "    nb.save(warped_img_nb, os.path.join(data_path_warped , warped_image_name + \"_\" + str(counter) + \".nii.gz\"))\n",
        "    counter = counter + 1\n",
        "  \n",
        "  return\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7HMOgagXvEmv"
      },
      "source": [
        "function that does below stuff\n",
        "- takes list of fixed images, moving images and deformation fields\n",
        "- uses respective optimizer, decay to train on an image pair for defined epochs\n",
        "- saves the warped image every few epochs"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U_J-4bwvuDu1"
      },
      "source": [
        "def direct_optimization_training(deformation_field_tnsr_lst = deformation_field_tnsr_lst, fixed_img_tnsr_lst = fixed_img_tnsr_lst,\n",
        "                                 moving_img_tnsr_lst = moving_img_tnsr_lst, optimizer = optimizer_dict['optimizer_adam'], \n",
        "                                 decay=scheduler_lst, epochs = 1001, decay_flag=False):\n",
        "  losses_dict = {}\n",
        "  counter = 0\n",
        "  for fixed_img_path, moving_img_path, fixed_img_tnsr, moving_img_tnsr, deformation_field in zip(data_path_lst_fixed, data_path_lst_moving,fixed_img_tnsr_lst, moving_img_tnsr_lst, deformation_field_tnsr_lst):\n",
        "    fixed_image_name = os.path.basename(fixed_img_path).replace(\".nii.gz\", \"\")\n",
        "    moving_image_name = os.path.basename(moving_img_path).replace(\".nii.gz\", \"\")\n",
        "    warped_image_name = fixed_image_name + \"|\" + moving_image_name\n",
        "\n",
        "    print(\"Optimizing {} image pair for {}, {}\".format(counter, fixed_image_name, moving_image_name ))\n",
        "\n",
        "    # obtain a list of deformation field at multiple of 500 epochs and also losses list at every multiple of 100th epoch\n",
        "    optimized_deformation_field_lst, loss_lst = optimize_one_image_pair(fixed_img_tnsr, moving_img_tnsr, deformation_field, \n",
        "                                                                        optimizer = optimizer[counter], scheduler = scheduler_lst[counter], \n",
        "                                                                        epochs=epochs, decay_flag=decay_flag)\n",
        "    \n",
        "    # saving the warped images for specific epoch deformation fields in the folder path\n",
        "    generate_and_save_warped_images(fixed_img_path, moving_img_path, fixed_img_tnsr, moving_img_tnsr, optimized_deformation_field_lst)\n",
        "\n",
        "    losses_dict[warped_image_name] = loss_lst\n",
        "    print(\"\")\n",
        "    print(\" ============ ================== ============== =============== ============ ================ ================ \")\n",
        "    print(\"\")\n",
        "    \n",
        "    counter = counter + 1\n",
        "\n",
        "    del optimized_deformation_field_lst\n",
        "  \n",
        "  return losses_dict\n",
        "    "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kNi7mT99uD0L",
        "outputId": "a115c696-15bd-4302-d35d-815ac919775b"
      },
      "source": [
        "losses_dict = direct_optimization_training(deformation_field_tnsr_lst = deformation_field_tnsr_lst, fixed_img_tnsr_lst = fixed_img_tnsr_lst,\n",
        "                                 moving_img_tnsr_lst = moving_img_tnsr_lst, optimizer = optimizer_dict['optimizer_rmsprop'], epochs = 1001)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Optimizing 0 image pair for IXI044-Guys-0712-T1, IXI044-Guys-0712-T2\n",
            "losses at 0 are tensor([1.7297], device='cuda:0', grad_fn=<MulBackward0>), 1.0003470274355664e-10, 0.0 and tensor([1.7297], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "losses at 500 are tensor([1.1198], device='cuda:0', grad_fn=<MulBackward0>), 0.0008038009400479496, 0.0 and tensor([1.1207], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "losses at 1000 are tensor([1.0793], device='cuda:0', grad_fn=<MulBackward0>), 0.0014493637718260288, 0.0 and tensor([1.0808], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "\n",
            " ============ ================== ============== =============== ============ ================ ================ \n",
            "\n",
            "Optimizing 1 image pair for IXI054-Guys-0707-T1, IXI054-Guys-0707-T2\n",
            "losses at 0 are tensor([1.7781], device='cuda:0', grad_fn=<MulBackward0>), 1.000449445509588e-10, 0.0 and tensor([1.7781], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "losses at 500 are tensor([1.2108], device='cuda:0', grad_fn=<MulBackward0>), 0.0006900625303387642, 0.0 and tensor([1.2115], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "losses at 1000 are tensor([1.1721], device='cuda:0', grad_fn=<MulBackward0>), 0.0012348524760454893, 0.0 and tensor([1.1734], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "\n",
            " ============ ================== ============== =============== ============ ================ ================ \n",
            "\n",
            "Optimizing 2 image pair for IXI061-Guys-0715-T1, IXI061-Guys-0715-T2\n",
            "losses at 0 are tensor([1.7770], device='cuda:0', grad_fn=<MulBackward0>), 9.998901012719941e-11, 0.0 and tensor([1.7770], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "losses at 500 are tensor([1.2503], device='cuda:0', grad_fn=<MulBackward0>), 0.0009345798753201962, 0.0 and tensor([1.2512], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "losses at 1000 are tensor([1.2080], device='cuda:0', grad_fn=<MulBackward0>), 0.0017848375719040632, 0.0 and tensor([1.2098], device='cuda:0', grad_fn=<AddBackward0>)\n",
            "\n",
            " ============ ================== ============== =============== ============ ================ ================ \n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BE0aKbWGJWn9"
      },
      "source": [
        "yy"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wHBBFxELPKV2"
      },
      "source": [
        "# Evaluation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y7UTLCRLRY9R"
      },
      "source": [
        "data_path_warped = '/content/drive/MyDrive/DATASET_FINAL/Test/intermodal/Batch_1/direct_optimization_sgd_T1_F_T2_M'\n",
        "data_path_lst_warped = sorted(glob.glob(os.path.join(data_path_warped, \"*.nii.gz\")))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZbpGsNPCV5lT"
      },
      "source": [
        "# warped_img_tnsr = spatial_transformer_deformable(moving_img_tnsr, deformation_field)\n",
        "# warped_image_np = warped_img_tnsr.detach().to(\"cpu\").numpy()\n",
        "# plt.imshow(warped_image_np[0, 0, :, :, 55], cmap='Greys_r')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "56PE5hEaPvX3"
      },
      "source": [
        "# moving_image_np = moving_img_tnsr.to(\"cpu\").numpy()\n",
        "# fixed_image_np = fixed_img_tnsr.to(\"cpu\").numpy()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "id": "aPsH7FWZ-5Kc",
        "outputId": "4191608e-c806-450f-c71d-cfa46ec9bb97"
      },
      "source": [
        "data_path_lst_moving[0]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'/content/drive/MyDrive/DATASET_FINAL/Test/intermodal/Batch_1/T1_F_T2_M/T2_affine/IXI044-Guys-0712-T2.nii.gz'"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7a1UBwQkR0P-"
      },
      "source": [
        "moving_image_np = run3functions(data_path_lst_moving[1])\n",
        "fixed_image_np = run3functions(data_path_lst_fixed[1])\n",
        "warped_image_np = run3functions(data_path_lst_warped[1])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "FtCNpQvzPMRg",
        "outputId": "eebf1e95-017f-4f22-940f-4e8a73d8f6d8"
      },
      "source": [
        "fig = plt.figure(figsize=(20, 15), dpi=80)\n",
        "\n",
        "#plt.subplot(1, 10, 1)\n",
        "ax1 = fig.add_subplot(131)\n",
        "ax1.title.set_text('Fixed')\n",
        "plt.imshow(fixed_image_np[0, 0, :, 51, :], cmap='Greys_r')\n",
        "\n",
        "#plt.subplot(1, 10, 2)\n",
        "\n",
        "ax2 = fig.add_subplot(132)\n",
        "ax2.title.set_text('Moving')\n",
        "plt.imshow(moving_image_np[0, 0, :, 51, :], cmap='Greys_r')\n",
        "\n",
        "\n",
        "ax3 = fig.add_subplot(133)\n",
        "ax3.title.set_text('Warped')\n",
        "#plt.subplot(1, 10, 3)\n",
        "plt.imshow(warped_image_np[0, 0, :, 51, :], cmap='Greys_r')\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1600x1200 with 3 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GO5XBBHLPMUZ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "72d1f9fb-9e8e-4212-b1e9-78b608ee909f"
      },
      "source": [
        "plt.imshow( warped_image_np[0, 0, :, 51, :]-moving_image_np[0, 0, :, 51, :] , cmap='Greys_r')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7faafce52750>"
            ]
          },
          "metadata": {},
          "execution_count": 48
        },
        {
          "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": "tuq7d58DMb0Y"
      },
      "source": [
        "warped_image_tnsr = spatial_transformer_deformable(deformable_moving_img_tnsr, deformable_tnsr)\n",
        "warped_image_np = warped_image_tnsr.detach().to(\"cpu\").numpy()\n",
        "plt.imshow(warped_image_np[0, 0, :, 75, :])"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}