Abstract
The new coronavirus (COVID-19) appeared in Wuhan in December 2019 and has been announced as a pandemic by the World Health Organization (WHO). Currently, this deadly pandemic has caused more than 1 million deaths worldwide. Therefore, it is essential to detect positive cases as early as possible to prevent the further spread of this outbreak. Currently, the most widely used COVID-19 detection technique is a real-time reverse transcription-polymerase chain reaction (RT-PCR). However, RT-PCR is time-consuming to confirm infection in the patient. Because RT-PCR is less sensitive, it provides high false-negative results. Computed tomography (CT) is recommended as a solution to this problem by healthcare professionals because of its higher sensitivity for early and rapid diagnosis. In addition, radiation used in CT poses a serious threat to patients. In this study, we propose a CNN-based method to distinguish COVID-19 pneumonia from other types of viral and bacterial pneumonia using low-dose CT images to reduce the radiation dose used in CT. In our study, we used a data set consisting of 7717 CT images of 350 patients that we collected from Çanakkale Onsekiz Mart University Research Hospital. We used a CNN-based network that suppresses noise to remove interference from low-dose CT images. In the image preprocessing phase, we provided lung segmentation from CT images and applied quantum Fourier transform. By evaluating all possible variations of local knowledge at the same time with quantum Fourier transformation, the most informative spatial information was extracted. In CNN-based architecture, we used pre-trained ResNet50v2 as a feature extractor and fine-tune by training with our dataset. We visualized the efficiency of the ResNet50v2 network using the t-SNE method. We performed the classification process with a fully connected layer. We created a heat map using the GradCam technique to see where the model focuses on the images while classifying. In this experimental study, the results of 99.5%, 99.2%, 99.0%, 99.7%, and 99.1%, were obtained in the context of performance criteria such as accuracy, precision, sensitivity, specificity, and f1 score, respectively. This study revealed the artificial intelligence-based computer-aided diagnosis (CAD)system as an effective and fast method to accurately diagnose COVID-19 pneumonia.
Competing Interest Statement
The authors have declared no competing interest.
Funding Statement
No payments or services have been received from a third party.
Author Declarations
I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.
Yes
The details of the IRB/oversight body that provided approval or exemption for the research described are given below:
2011-KAEk-27/2020-E.2000060082 number of project has been given ethical committee approval by Clinical Research Ethics Committee of Canakkale Onsekiz Mart University of Republic of Turkey
All necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.
Yes
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
Yes
I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.
Yes
Footnotes
E-mail: erdiacar{at}stu.comu.edu.tr
E-mail: bilgeoztoprak{at}gmail.com
Data Availability
The data set used in this study was collected from Canakkale Onsekiz Mart University Research Hospital with ethical permissions.
