tumorseg.py

import os
import tensorflow as tf
from tensorflow.keras import backend as K
import tf_keras
import numpy as np
import matplotlib.pyplot as plt
import matplotlib
matplotlib.use('agg')
def dice_coefficients(y_true, y_pred, smooth=100):
    y_true_flatten = K.flatten(y_true)
    y_pred_flatten = K.flatten(y_pred)

    intersection = K.sum(y_true_flatten * y_pred_flatten)
    union = K.sum(y_true_flatten) + K.sum(y_pred_flatten)
    return (2 * intersection + smooth) / (union + smooth)


def dice_coefficients_loss(y_true, y_pred, smooth=100):
    return 1.0 - dice_coefficients(y_true, y_pred, smooth)


def iou(y_true, y_pred, smooth=100):
    intersection = K.sum(y_true * y_pred)
    sum = K.sum(y_true + y_pred)
    iou = (intersection + smooth) / (sum - intersection + smooth)
    return iou


def jaccard_distance(y_true, y_pred):
    y_true_flatten = K.flatten(y_true)
    y_pred_flatten = K.flatten(y_pred)
    return -iou(y_true_flatten, y_pred_flatten)

segmodel = tf_keras.models.load_model("segment_model/V2", custom_objects={'dice_coefficients_loss': dice_coefficients_loss, 'iou': iou, 'dice_coefficients': dice_coefficients  } )

def load_image_for_pred(image_path):
    img = tf.keras.utils.load_img(
        image_path,
        color_mode='rgb',
        target_size=(256, 256),
        interpolation='nearest',
        keep_aspect_ratio=False
    )
    img = tf.keras.utils.img_to_array(img) / 255
    return np.array([img])

def make_segmentation(image_path):
    img = load_image_for_pred(image_path)
    predicted_img = segmodel.predict(img)

    plt.figure(figsize=(5, 3))
    plt.subplot(1, 3, 1)
    plt.imshow(np.squeeze(img))
    plt.title('Original Image')
    plt.axis(False)


    plt.subplot(1, 3, 2)
    plt.imshow(np.squeeze(predicted_img) > 0.5)
    plt.title('Prediction')
    plt.axis(False)

    plt.subplot(1, 4, 4)
    plt.imshow(np.squeeze(img))
    plt.imshow(np.squeeze(predicted_img) > 0.5, cmap='gray', alpha=0.5)
    plt.title('Image w/h Mask')
    plt.axis(False)
    save_file_name = os.path.splitext(image_path)[0] + '_segmented.png'
    plt.savefig(save_file_name)

    return save_file_name