estudio_redes.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "from tensorflow import keras\n",
    "import tensorflow as tf\n",
    "import random\n",
    "import matplotlib.pyplot as plt\n",
    "import unets_repo\n",
    "from keras import backend as K\n",
    "from scipy.stats import kurtosis, skew\n",
    "from skimage.measure import shannon_entropy\n",
    "import ast"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "fp1 = r\"C:\\Users\\Pier\\Desktop\\PFI_main\\DS_VOLUMEN\\main\\train_data\"\n",
    "fp2 = r\"C:\\Users\\Pier\\Desktop\\PFI_main\\DS_VOLUMEN\\main\\validation_data\"\n",
    "\n",
    "def load_data(ids,suj_path,mask_path):\n",
    "  suj_files = os.listdir(suj_path)\n",
    "  mask_files = os.listdir(mask_path)\n",
    "  X = []\n",
    "  Y = []\n",
    "\n",
    "  for i in ids:\n",
    "    x = np.load(os.path.join(suj_path,suj_files[i]))\n",
    "    y = np.load(os.path.join(mask_path,mask_files[i]))\n",
    "\n",
    "    X.append(x)\n",
    "    Y.append(y)\n",
    "\n",
    "  return np.array(X), np.array(Y)\n",
    "def batch_generator(suj_path, mask_path):\n",
    "    suj_files = os.listdir(suj_path)\n",
    "    indices = np.arange(len(suj_files))\n",
    "    indices = np.random.permutation(indices)\n",
    "    batch=[]\n",
    "    while True:\n",
    "            for i in indices:\n",
    "                batch.append(i)\n",
    "                if len(batch)==batch_size:\n",
    "                    yield load_data(batch,suj_path,mask_path)\n",
    "                    batch=[]\n",
    "\n",
    "train_gen_ds = batch_generator(os.path.join(fp1,\"sujetos\"),os.path.join(fp1,\"mascaras\"))\n",
    "val_gen_ds = batch_generator(os.path.join(fp2,\"sujetos\"),os.path.join(fp2,\"mascaras\"))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Métricas con casteo a Double para calcular hard dice score a mano"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Dice(targets, inputs):    \n",
    "       \n",
    "    inputs = K.flatten(inputs)\n",
    "    targets = K.flatten(targets)\n",
    "    inputs = tf.cast(inputs, dtype=tf.double)\n",
    "    targets = tf.cast(targets, dtype=tf.double)\n",
    "    intersection = K.sum(targets * inputs)\n",
    "    \n",
    "    return (2*intersection + K.epsilon()) / (K.sum(targets) + K.sum(inputs) + K.epsilon())\n",
    "\n",
    "def Jaccard(targets, inputs):\n",
    "    \n",
    "    inputs = K.flatten(inputs)\n",
    "    targets = K.flatten(targets)\n",
    "    inputs = tf.cast(inputs, dtype=tf.double)\n",
    "    targets = tf.cast(targets, dtype=tf.double)\n",
    "    \n",
    "    intersection = K.sum(K.abs(inputs * targets))\n",
    "    union = K.sum(inputs) + K.sum(targets) - intersection\n",
    "    \n",
    "    return (intersection + K.epsilon())/(union + K.epsilon())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 1\n",
    "n_val_img = len(os.listdir(os.path.join(fp2,\"sujetos\")))\n",
    "vspe = n_val_img//batch_size"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Todos los modelos analizados"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "attunetselu = unets_repo.AttUnet(input_size=(256,256,1),activation=\"selu\",initializer=\"lecun_normal\",num_filters=64)\n",
    "attunetselu.load_weights(\"modelsh5/0.8160_attunet_selu_lecun_aug_dicebce.h5\")\n",
    "unet3 = unets_repo.unet3plus(input_size=(256,256,1))\n",
    "unet3.load_weights(\"modelsh5/0.5662_unet3plus_relu_aug_dicebce.h5\")\n",
    "vsunet = unets_repo.UNet(input_size=(256,256,1),activation=\"selu\",initializer=\"lecun_normal\",num_filters=16)\n",
    "vsunet.load_weights(\"modelsh5/0.8174_unet_v_small_selu_lecun_aug_dice.h5\")\n",
    "sunet = unets_repo.UNet(input_size=(256,256,1),activation=\"relu\",initializer=\"he_uniform\",num_filters=32)\n",
    "sunet.load_weights(\"modelsh5/0.7947_unet_small_relu_he_aug.h5\")\n",
    "unet = unets_repo.UNet(input_size=(256,256,1),activation=\"relu\",initializer=\"he_uniform\",num_filters=64)\n",
    "unet.load_weights(\"modelsh5/0.8134_unet_relu_he_aug.h5\")\n",
    "unetselu = unets_repo.UNet(input_size=(256,256,1),activation=\"selu\",initializer=\"lecun_normal\",num_filters=64)\n",
    "unetselu.load_weights(\"modelsh5/0.8314_unet_selu_lecun_aug_dicebce.h5\")\n",
    "restattunet = unets_repo.ResAttUnet(input_size=(256,256,1),activation=\"selu\",initializer=\"lecun_normal\",num_filters=64,bn=True)\n",
    "restattunet.load_weights(\"modelsh5/0.8026_resattunet_selu_bn_dicebce.h5\")\n",
    "classicunet = unets_repo.classicUNet(input_size=(256,256,1),activation=\"selu\",initializer=\"lecun_normal\",num_filters=64)\n",
    "classicunet.load_weights(\"modelsh5/0.8233_classicunet_selu_lecun_aug_dicebce.h5\")\n",
    "attunet = unets_repo.AttUnet(input_size=(256,256,1),activation=\"relu\",initializer=\"he_uniform\",num_filters=64)\n",
    "attunet.load_weights(\"modelsh5/0.8042_attunet_relu_aug_dicebce.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "36986313\n"
     ]
    }
   ],
   "source": [
    "print(attunetselu.count_params())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "nets = [attunetselu,unet3,vsunet,sunet,unet,unetselu,restattunet,classicunet,attunet]\n",
    "netnames = [\"AttUnet w/s\",\"Unet3+\",\"vsUnet\",\"sUnet\",\"Unet\",\"Unet w/s\",\"ResAttUnet\",\"ClassicUnet\",\"AttUnet\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AttUnet w/s 36986313\n",
      "Unet3+ 27719685\n",
      "vsUnet 1946709\n",
      "sUnet 7771301\n",
      "Unet 31054149\n",
      "Unet w/s 31054149\n",
      "ResAttUnet 38393737\n",
      "ClassicUnet 31054149\n",
      "AttUnet 36986313\n"
     ]
    }
   ],
   "source": [
    "for i,net in enumerate(nets):\n",
    "    print(netnames[i],net.count_params())"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Búsqueda de Thresholds:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dices = []\n",
    "for net in nets:\n",
    "    dice = []\n",
    "    for th in [0.05+0.05*i for i in range(19)]: #0.05-1\n",
    "        dice_th = 0\n",
    "        for _ in range(n_val_img):\n",
    "            test_image_batch, test_mask_batch = val_gen_ds.__next__()\n",
    "            test_img = test_image_batch[0]\n",
    "            ground_truth = test_mask_batch[0]\n",
    "            test_img_norm = test_img[:,:]\n",
    "            test_img_input = np.expand_dims(test_img_norm, 0)\n",
    "            prediction = nets[0](test_img_input).numpy()\n",
    "            predicted_img_th = (prediction[0,:,:,0]>th)*1\n",
    "            dice_th += Dice(ground_truth,predicted_img_th).numpy()\n",
    "        dice.append(dice_th/n_val_img)\n",
    "    dices.append(dice)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('dices_by_th.txt', 'r') as file:\n",
    "    dices = ast.literal_eval(file.read())\n",
    "dices = [dice[0:-2] for dice in dices]\n",
    "maxs = [max(d) for d in dices]\n",
    "indices = [n.index(maxs[i])*0.05+0.05 for i,n in enumerate(dices)]\n",
    "performance_modelos = dict(zip(netnames,maxs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(6,6))\n",
    "for i,n in enumerate(dices):\n",
    "    plt.plot(n,label=netnames[i])\n",
    "    plt.xticks([0,5,10,15,19],[0.05,0.30,0.55,0.80,1.0])\n",
    "    plt.legend()\n",
    "    plt.xlabel(\"Threshold\")\n",
    "    plt.ylabel(\"Dice Score\")\n",
    "    #plt.title(\"p/set Validación\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "#coeff =   [95, 123, 52, 131, 64, 76, 66, 89, 112] #cuantas fueron el mejor dice de las imagenes con prostata\n",
    "#          [s,  r,   s,  r,   r,  s,  s,  s,  r]\n",
    "       #  [90, 120, 47, 126, 60, 74, 62, 87, 107, 35]\n",
    "#coeff_n = [559, 512, 554, 553, 556, 532, 537, 544, 545] # +1 si dijieron q es full background y era.\n",
    "indices = [0.95, 0.45, 0.7, 0.65, 0.85, 0.95, 0.95, 0.95, 0.75]\n",
    "#pond = np.add(coeff,coeff_n)\n",
    "\n",
    "#todas nets = 0.859 th 0.5\n",
    "#att unet x2 + unet3+ small = 0.859 th 0.4\n",
    "#att w/s + unet3 + small = 0.857 th 0.5\n",
    "#att w/s + unet3 + small + classic = 0.86 th 0.4"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Búsqueda de coeficientes / performance por modelo:"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Análisis sobre imágenes SIN próstata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "scores = [0]*len(nets2)\n",
    "for _ in range(vspe):\n",
    "    test_image_batch, test_mask_batch = val_gen_ds.__next__()\n",
    "    test_img = test_image_batch[0]\n",
    "    ground_truth = test_mask_batch[0]\n",
    "    test_img_norm = test_img[:,:]\n",
    "    test_img_input = np.expand_dims(test_img_norm, 0)\n",
    "    dice = 0\n",
    "    if not np.any(ground_truth):\n",
    "        for i,n in enumerate(nets):\n",
    "            prediction = n(test_img_input).numpy()\n",
    "            predicted_img_th = (prediction[0,:,:,0]>indices[i])*1\n",
    "            dd = Dice(ground_truth,predicted_img_th).numpy()\n",
    "            if dd == 1:\n",
    "                scores[i]+=1\n",
    "print(scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "scores3 = [0]*len(nets2)\n",
    "for _ in range(vspe):\n",
    "    test_image_batch, test_mask_batch = val_gen_ds.__next__()\n",
    "    test_img = test_image_batch[0]\n",
    "    ground_truth = test_mask_batch[0]\n",
    "    test_img_norm = test_img[:,:]\n",
    "    test_img_input = np.expand_dims(test_img_norm, 0)\n",
    "    dice = 0\n",
    "    for i,n in enumerate(nets2):\n",
    "        prediction = n(test_img_input).numpy()\n",
    "        predicted_img_th = (prediction[0,:,:,0]>index[i])*1\n",
    "        dd = Dice(ground_truth,predicted_img_th).numpy()\n",
    "        if ((dd!=0) and (dd!=1)):\n",
    "            if dd>dice:\n",
    "                dice = dd\n",
    "                max_i = i\n",
    "    if dice > 0:\n",
    "        scores3[max_i]+=1\n",
    "print(scores3)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### En conclusión quedaron 4 modelos:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "nets2 = [nets[0],nets[1],nets[3],nets[7]]\n",
    "index = [indices[0],indices[1],indices[3],indices[7]]\n",
    "coeff2 = [169, 201, 228, 161] #0137"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dice3 = 0\n",
    "for _ in range(n_val_img):\n",
    "    test_image_batch, test_mask_batch = val_gen_ds.__next__()\n",
    "    test_img = test_image_batch[0]\n",
    "    ground_truth = test_mask_batch[0]\n",
    "    test_img_norm = test_img[:,:]\n",
    "    test_img_input = np.expand_dims(test_img_norm, 0)\n",
    "    prediction = np.zeros(shape=(1,256,256,256))\n",
    "    for i,net in enumerate(nets):\n",
    "        prediction += ((net(test_img_input).numpy())>index[i])*coeff2[i]/sum(coeff2)\n",
    "    predicted_img_th = (prediction[0,:,:,0]>0.4)*1\n",
    "    dice3 += Dice(ground_truth,predicted_img_th).numpy()\n",
    "print(dice3/n_val_img) #0.86"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x800 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_image_batch, test_mask_batch = val_gen_ds.__next__()\n",
    "test_img = test_image_batch[0]\n",
    "ground_truth = test_mask_batch[0]\n",
    "test_img_norm = test_img[:,:]\n",
    "test_img_input = np.expand_dims(test_img_norm, 0)\n",
    "\n",
    "plt.figure(figsize=(15,8))\n",
    "plt.subplot(341)\n",
    "plt.title('Testing Image')\n",
    "plt.imshow(test_img, cmap='gray')\n",
    "plt.axis(\"off\")\n",
    "plt.subplot(342)\n",
    "plt.title('Testing Label')\n",
    "plt.imshow(ground_truth>0.5, cmap='gray')\n",
    "plt.axis(\"off\")\n",
    "images = []\n",
    "for i,n in enumerate(nets):\n",
    "    prediction = n(test_img_input).numpy()\n",
    "    prediction_th= (prediction[0,:,:,0]>indices[i])*1\n",
    "    dice = Dice(ground_truth,prediction_th).numpy()\n",
    "    plt.subplot(3,4,i+3)\n",
    "    plt.title(netnames[i]+\" - \"+str(round(dice,3)))\n",
    "    plt.imshow(prediction_th, cmap='gray')\n",
    "    plt.axis(\"off\")\n",
    "    #images.append((prediction[0,:,:,0]>indices[i])*1)\n",
    "prediction = np.zeros(shape=(1,256,256,256))\n",
    "for i,net in enumerate(nets2):\n",
    "    prediction += ((net(test_img_input).numpy())>index[i])*coeff2[i]/sum(coeff2)\n",
    "predicted_img_th = (prediction[0,:,:,0]>0.4)*1\n",
    "dice = Dice(ground_truth,predicted_img_th).numpy()\n",
    "plt.subplot(3,4,12)\n",
    "plt.title(\"MoE - \"+str(round(dice,3)))\n",
    "plt.imshow(predicted_img_th, cmap='gray')\n",
    "plt.axis(\"off\")\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Búsqueda de TH dinámico a partir de características de imágenes de entrada\n",
    "> Se descarta, media, std, sum(pixeles), histogram related, kurtosis, entropía, skewness, correlacion 2d. <br>\n",
    "> No hay resultados viables, se propone usar threshold según dice score en validation set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "netnames = [\"AttUnet w/s\",\"Unet3+\",\"sUnet\",\"ClassicUnet\"]\n",
    "nets = [attunetselu,unet3,sunet,classicunet]\n",
    "thresholds = [0.95,0.45,0.65,0.95] # th talque hard dice score es máx en set de validación\n",
    "coeff = [169, 201, 228, 161] # cantidad de imágenes CON prostata donde obtuvieron mejor hard dice score\n",
    "moe_th = 0.4 # th talque hard dice score es máx en set de validación\n",
    "for unique in np.unique(prediction):\n",
    "    test_img = test_image_batch[0]\n",
    "    ground_truth = test_mask_batch[0]\n",
    "    test_img_norm = test_img[:,:]\n",
    "    test_img_input = np.expand_dims(test_img_norm, 0)\n",
    "    prediction = np.zeros(shape=(1,256,256,256))\n",
    "    for i,net in enumerate(nets):\n",
    "        prediction += ((net(test_img_input).numpy()) > thresholds[i])*coeff[i]/sum(coeff)\n",
    "    predicted_img_th = (prediction[0,:,:,0] > unique)*1\n",
    "    print(unique, Dice(ground_truth,predicted_img_th).numpy())\n",
    "test_img_ravel = np.ravel(test_img)\n",
    "cor = np.corrcoef(test_img)\n",
    "cor_ravel = np.ravel(cor)\n",
    "plt.hist(test_img_ravel,bins=1000)\n",
    "plt.show()\n",
    "print(np.sum(test_img_ravel))\n",
    "print(kurtosis(test_img_ravel))\n",
    "print(shannon_entropy(test_img))\n",
    "print(skew(test_img_ravel))\n",
    "plt.imshow(cor)\n",
    "plt.show()\n",
    "print(np.sum(cor_ravel))\n",
    "print(kurtosis(cor_ravel))\n",
    "print(shannon_entropy(cor))\n",
    "print(skew(cor_ravel))"
   ]
  }
 ],
 "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": "3.8.0"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "a8e68cbbe928a735cbef36e0c504d1e4907a06cf8776305a5ea24ea01ddf5907"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}