A Fully Automated Deep Learning-based Network For Detecting COVID-19 from a New And Large Lung CT Scan Dataset

Abstract

<jats:p>COVID-19 is a severe global problem, and one of the primary ways to decrease its casualties is the infected person's identification at the proper time. AI can play a significant role in these cases by monitoring and detecting infected persons in early-stage. In this paper, we aim to propose a high- speed and accurate fully-automated method to detect COVID-19 from the patient's CT scan. We introduce a new dataset that contains 48260 CT scan images from 282 normal persons and 15589 images from 95 patients with COVID-19 infection. Our proposed automated system takes all the CT scan image sequences of a patient as the input and determines if the patient is infected with COVID-19. At the first stage, this system runs the proposed image processing algorithm to discard those CT images that inside the lung is not properly visible in them. This helps to reduce the number of images that shall be processed, so it reduces the processing time. Also, running this algorithm makes the deep network at the next stage to analyze only the proper images and thus reduces false detections. At the next stage, we propose a new modified deep convolutional network that is based on ResNet50V2 and is enhanced by the feature pyramid network for classifying the selected CT images into COVID-19 or normal. After running these two phases, if enough number of chosen CT scan images of a patient be identified as COVID-19, the system considers that patient, infected to this disease. In the single image classification stage, the ResNet50V2 with feature pyramid network achieved 98.49% accuracy on more than 7996 validation images. At the fully automated phase, the automated system correctly identified almost 237 patients from 245 patients on average between five-folds with high speed. In the end, we also investigate the classified images with a feature visualization algorithm to indicate the area of infections in each image. We are implementing these materials on some medical centers in Iran, and we hope that it would be a great help in Intelligence disease detection anywhere.</jats:p>

A deep learning algorithm to detect coronavirus (COVID-19) disease using CT images

“…Full-size  DOI: 10.7717/peerj-cs.345/fig- 1 preprocessing stages are applied on images to select special images of a breath sequence or highlight lung infected area, before entering them to the classification algorithm (Rahimzadeh, Attar & Sakhaei, 2020). In order to have a fully automated algorithm, in this paper no preprocessing, preselecting or ROI selecting is performed on the images.…”

“…The data used in this paper is downloaded from publicly available dataset (Rahimzadeh, Attar & Sakhaei, 2020). They have collected 15,589 CT images of 95 positive patients and 48,260 images of 282 negative persons.…”

“…Data are available in Rahimzadeh, Attar, and Sakhaei (2020) and at GitHub: https://github.com/mr7495/COVID-CTset.…”

Abnormality detection and intelligent severity assessment of human chest computed tomography scans using deep learning: a case study on SARS-COV-2 assessment

“…Open lung slices refer to slices with lung parenchyma, while closed lung slices contain mainly bones. The reason of conversion to 2D slices is to efficiently select the correct candidate slices with infections from all slices in 3D image sequence (Hamadi and Yagoub 2018;Rahimzadeh et al 2020). Slice selection is used to select the informative slices (open lung slices) and reject the remaining slices, which shall positively affect training time, model accuracy and precision to generate an efficient classification model (Hamadi and Yagoub 2018;Rahimzadeh et al 2020).…”

“…The reason of conversion to 2D slices is to efficiently select the correct candidate slices with infections from all slices in 3D image sequence (Hamadi and Yagoub 2018;Rahimzadeh et al 2020). Slice selection is used to select the informative slices (open lung slices) and reject the remaining slices, which shall positively affect training time, model accuracy and precision to generate an efficient classification model (Hamadi and Yagoub 2018;Rahimzadeh et al 2020). An automatic slice selection technique is needed to speed up slice selection stage to save a lot of time and effort in comparison to manually selecting the desired open lung slices based on medical expert decisions (Rahimzadeh et al 2020).…”