CNN supported automated recognition of Covid-19 infection in chest X-ray images

Abstract

Automatic recognition of lung system is use to identify normal and covid infected lungs from chest X-ray images of the people. In the year 2020, the coronavirus forcefully pushed the entire world into a freakish situation, the foremost challenge is to diagnosis the coronavirus. We have got standard diagnosis test called PCR test which is complex and costlier to check the patient's sample at initial stage. Keeping this in mind, we developed a work to recognize the chest X-ray image automatically and label it as Covid or normal lungs. For this work, we collected the dataset from open-source data repository and then pre-process each X-ray images from each category such as covid X-ray images and non-covid X-ray images using various techniques such as filtering, edge detection, segmentation, etc., and then the pre-processed X-ray images are trained using CNN-Resnet18 network. Using PyTorch python package, the resnet-18 network layer is created which gives more accuracy than any other algorithm. From the acquired knowledge the model is correctly classifies the testing X-ray images. Then the performance of the model is calculated and analyzed with various algorithms and hence gives that the resnet-18 network improves our model performance in terms of specificity and sensitivity with more than 90%.

Keywords: Canny algorithm; Chest X-ray; PyTorch; Resnet18.

Fig. 1

Entire workflow of the lung recognition model.

Fig. 2

Raw dataset images are collected from open-source data repository nearly 1500 X-ray images of both covid and non-covid lungs for the further process.

Fig. 3

Pre-processing the input data from each category one image is taken to perform various per-processing technique for example, like this all the image will be pre-processed before training the image to the model. The X-Ray image(1.a and 2.a) is taken to remove the unwanted noise from the image using median filter(1.b and 2.b), then edges of the image is detected using Canny algorithm(1.c and 2.c) and segmented using binary threshold(1.d and 2.d) and finally the image is contour(1.e and 2.e) for the better understanding of the image.

Fig. 3

Pre-processing the input data from each category one image is taken to perform various per-processing technique for example, like this all the image will be pre-processed before training the image to the model. The X-Ray image(1.a and 2.a) is taken to remove the unwanted noise from the image using median filter(1.b and 2.b), then edges of the image is detected using Canny algorithm(1.c and 2.c) and segmented using binary threshold(1.d and 2.d) and finally the image is contour(1.e and 2.e) for the better understanding of the image.

Fig. 4

Basic and Standard structure of Res Net18 Architecture.

Fig. 5

Detailed view of Single Residual Function.

Fig. 6

Testing the image after training the model.

Fig. 7

Accuracy and loss of the ResNet18 model during validation.