Neural Network for Pneumonia Detection Chest X-Ray Images¶

In [5]:

import tensorflow as tf
import numpy as np
import os
import sys
import cv2
import matplotlib.pyplot as plt
import pickle
import random
import pandas as pd

import tensorflow as tf
from tensorflow.keras import Sequential
from tensorflow.keras.layers import Dense,Dropout,Activation, Flatten,Conv2D,MaxPooling2D
import pickle

from sklearn.model_selection import train_test_split

import matplotlib.pyplot as plt
%matplotlib inline

In [3]:

class MasterImage(object):

    def __init__(self,PATH='', IMAGE_SIZE = 50, CATEGORIES=[]):
        self.PATH = PATH
        self.IMAGE_SIZE = IMAGE_SIZE
        self.CATEGORIES = CATEGORIES
        self.image_data = []
        self.x_data = []
        self.y_data = []

    def Process_Image(self):

        """
        Return Numpy array of image
        :return: X_Data, Y_Data
        """

        for categories in self.CATEGORIES:                                                  # Iterate over categories

            train_folder_path = os.path.join(self.PATH, categories)                         # Folder Path
            class_index = self.CATEGORIES.index(categories)                                 # this will get index for classification

            for img in os.listdir(train_folder_path):                                       # This will iterate in the Folder
                new_path = os.path.join(train_folder_path, img)                             # image Path

                try:        # if any image is corrupted
                    image_data_temp = cv2.imread(new_path,cv2.IMREAD_GRAYSCALE)                 # Read Image as numbers
                    image_temp_resize = cv2.resize(image_data_temp,(self.IMAGE_SIZE,self.IMAGE_SIZE))
                    self.image_data.append([image_temp_resize,class_index])
                except:
                    pass

        data = np.asanyarray(self.image_data)

        # Iterate over the Data
        for x in data:
            self.x_data.append(x[0])        # Get the X_Data
            self.y_data.append(x[1])        # get the label

        X_Data = np.asarray(self.x_data) / (255.0)      # Normalize Data
        Y_Data = np.asarray(self.y_data)

        return X_Data,Y_Data

    def pickle_image(self):

        """
        :return: None Creates a Pickle Object of DataSet
        """

        X_Data,Y_Data = self.Process_Image()

        pickle_out = open('X_Data','wb')
        pickle.dump(X_Data, pickle_out)
        pickle_out.close()

        pickle_out = open('Y_Data', 'wb')
        pickle.dump(Y_Data, pickle_out)
        pickle_out.close()

        print("Pickled Image Successfully ")

        return X_Data,Y_Data

    def load_dataset(self):

        try:
            X_Temp = open('X_Data','rb')
            X_Data = pickle.load(X_Temp)

            Y_Temp = open('Y_Data','rb')
            Y_Data = pickle.load(Y_Temp)

            print('Reading Dataset from PIckle Object')

            return X_Data,Y_Data

        except:
            print('Could not Found Pickle File ')
            print('Loading File and Dataset  ..........')
            X_Data,Y_Data = self.pickle_image()
            return X_Data,Y_Data

In [10]:

a = MasterImage(PATH='/Users/soumilshah/IdeaProjects/mytensorflow/Dataset/chest_xray/train',
                    IMAGE_SIZE=80,
                    CATEGORIES=['NORMAL','PNEUMONIA'])

X_Data,Y_Data = a.load_dataset()
X_Data = X_Data.reshape(-1,80,80,1)
print(X_Data.shape)

Reading Dataset from PIckle Object
(5216, 80, 80, 1)

In [11]:

X_Train, X_Test, Y_Train, Y_Test = train_test_split(X_Data, Y_Data, test_size=0.3,random_state=101)

In [12]:

model = Sequential()
model.add(Conv2D(200, (3, 3), input_shape=X_Data.shape[1:]))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Conv2D(100, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Flatten())
model.add(Dense(80))
model.add(Dense(1))
model.add(Activation('sigmoid'))
model.compile(loss='binary_crossentropy',
              optimizer='adam',
              metrics=['accuracy'])

WARNING:tensorflow:From /anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/resource_variable_ops.py:435: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.
Instructions for updating:
Colocations handled automatically by placer.

In [13]:

model.fit(X_Data, Y_Data, batch_size=40, epochs=3, validation_split=0.3)
model.save('doc.model')

Train on 3651 samples, validate on 1565 samples
WARNING:tensorflow:From /anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.cast instead.
Epoch 1/3
3651/3651 [==============================] - 217s 59ms/sample - loss: 0.3099 - acc: 0.8631 - val_loss: 0.0498 - val_acc: 0.9834
Epoch 2/3
3651/3651 [==============================] - 228s 62ms/sample - loss: 0.1117 - acc: 0.9548 - val_loss: 0.1889 - val_acc: 0.9291
Epoch 3/3
3651/3651 [==============================] - 218s 60ms/sample - loss: 0.1078 - acc: 0.9581 - val_loss: 0.0643 - val_acc: 0.9796

In [50]:

model.evaluate(X_Test,Y_Test,batch_size=40)

1565/1565 [==============================] - 18s 11ms/sample - loss: 0.0688 - acc: 0.9764

Out[50]:

[0.0687599335544216, 0.9763578]

In [ ]:

Test the DataSet¶

In [20]:

def prepare(filepath):
    training_date = []
    
    img_array = cv2.imread(filepath,cv2.IMREAD_GRAYSCALE)
    new_array = cv2.resize(img_array,(80,80))
    new_image =  new_array.reshape(-1,80,80,1)
    return new_image

In [21]:

model = tf.keras.models.load_model('doc.model')

In [29]:

filepath = '/Users/soumilshah/Downloads/NORMAL2-IM-1427-0001.jpeg'
img_array = cv2.imread(filepath)

plt.imshow(img_array)

Out[29]:

<matplotlib.image.AxesImage at 0xb38f69710>

Lets See what our Network has to say¶

In [30]:

test = model.predict([prepare(filepath=filepath)])

In [31]:

CATEGORIES=['NORMAL','PNEUMONIA']

In [41]:

print(CATEGORIES[int(test[0][0])])

PNEUMONIA

Great Lets Try Normal case¶

In [44]:

filepath = '/Users/soumilshah/Downloads/n.jpeg'
img_array = cv2.imread(filepath)

plt.imshow(img_array)

Out[44]:

<matplotlib.image.AxesImage at 0xb3fc0be80>

In [48]:

test = model.predict([prepare(filepath=filepath)])
CATEGORIES=['NORMAL','PNEUMONIA']
print(CATEGORIES[int(test[0][0])])

PNEUMONIA