AI can be nearly as accurate as a physician in diagnosing COVID-19 in the lungs


A University of Central Florida researcher is part of a new study showing that artificial intelligence can be nearly as accurate as a physician in diagnosing COVID-19 in the lungs.

The study, recently published in Nature Communications, shows the new technique can also overcome some of the challenges of current testing.

Researchers demonstrated that an AI algorithm could be trained to classify COVID-19 pneumonia in computed tomography (CT) scans with up to 90 percent accuracy, as well as correctly identify positive cases 84 percent of the time and negative cases 93 percent of the time.

CT scans offer a deeper insight into COVID-19 diagnosis and progression as compared to the often-used reverse transcription-polymerase chain reaction, or RT-PCR, tests. These tests have high false negative rates, delays in processing and other challenges.

Another benefit to CT scans is that they can detect COVID-19 in people without symptoms, in those who have early symptoms, during the height of the disease and after symptoms resolve.

However, CT is not always recommended as a diagnostic tool for COVID-19 because the disease often looks similar to influenza-associated pneumonias on the scans.

The new UCF co-developed algorithm can overcome this problem by accurately identifying COVID-19 cases, as well as distinguishing them from influenza, thus serving as a great potential aid for physicians, says Ulas Bagci, an assistant professor in UCF’s Department of Computer Science.

Bagci was a co-author of the study and helped lead the research.

“We demonstrated that a deep learning-based AI approach can serve as a standardized and objective tool to assist healthcare systems as well as patients,” Bagci says.

“It can be used as a complementary test tool in very specific limited populations, and it can be used rapidly and at large scale in the unfortunate event of a recurrent outbreak.”

Bagci is an expert in developing AI to assist physicians, including using it to detect pancreatic and lung cancers in CT scans.

He also has two large, National Institutes of Health grants exploring these topics, including $2.5 million for using deep learning to examine pancreatic cystic tumors and more than $2 million to study the use of artificial intelligence for lung cancer screening and diagnosis.

To perform the study, the researchers trained a computer algorithm to recognize COVID-19 in lung CT scans of 1,280 multinational patients from China, Japan and Italy.

Then they tested the algorithm on CT scans of 1,337 patients with lung diseases ranging from COVID-19 to cancer and non-COVID pneumonia.

When they compared the computer’s diagnoses with ones confirmed by physicians, they found that the algorithm was extremely proficient in accurately diagnosing COVID-19 pneumonia in the lungs and distinguishing it from other diseases, especially when examining CT scans in the early stages of disease progression.

“We showed that robust AI models can achieve up to 90 percent accuracy in independent test populations, maintain high specificity in non-COVID-19 related pneumonias, and demonstrate sufficient generalizability to unseen patient populations and centers,” Bagci says.

The UCF researcher is a longtime collaborator with study co-authors Baris Turkbey and Bradford J. Wood. Turkbey is an associate research physician at the NIH’s National Cancer Institute Molecular Imaging Branch, and Wood is the director of NIH’s Center for Interventional Oncology and chief of interventional radiology with NIH’s Clinical Center.

In this worldwide health crisis, the medical industry is looking for new technologies to monitor and controls the spread of COVID-19 (Coronavirus) pandemic.

AI is one of such technology which can easily track the spread of this virus, identifies the high-risk patients, and is useful in controlling this infection in real-time. It can also predict mortality risk by adequately analyzing the previous data of the patients. AI can help us to fight this virus by population screening, medical help, notification, and suggestions about the infection control [[1], [2], [3]].

This technology has the potential to improve the planning, treatment and reported outcomes of the COVID-19 patient, being an evidence-based medical tool. Fig. 1 shows the general procedure of AI and non-AI based applications that help general physicians to identify the COVID-19 symptoms.

Fig. 1
Fig. 1. General procedure of AI and non-AI based applications that help general physicians to identify the COVID-19 symptoms.

The above flow diagram informs and compares the flow of minimal non-AI treatment versus AI-based treatment. The above flow diagram explains the involvement of AI in the significant steps of treatment of high accuracy and reduces complexity and time taken. The physician is not only focused on the treatment of the patient, but also the control of disease with the AI application. Major symptoms and test analysis are done with the help of AI with the highest of accuracy. It also shows it reduces the total number of steps taken in the whole process, making more procurable in nature.

Main applications of AI in COVID-19 pandemic

I) Early detection and diagnosis of the infection

AI can quickly analyze irregular symptom and other ‘red flags’ and thus alarm the patients and the healthcare authorities [4,5]. It helps to provide faster decision making, which is cost-effective. It helps to develop a new diagnosis and management system for the COVID 19 cases, through useful algorithms. AI is helpful in the diagnosis of the infected cases with the help of medical imaging technologies like Computed tomography (CT), Magnetic resonance imaging (MRI) scan of human body parts.

II) Monitoring the treatment

AI can build an intelligent platform for automatic monitoring and prediction of the spread of this virus. A neural network can also be developed to extract the visual features of this disease, and this would help in proper monitoring and treatment of the affected individuals [[6], [7], [8]]. It has the capability of providing day-to-day updates of the patients and also to provide solutions to be followed in COVID-19 pandemic.

III) Contact tracing of the individuals

AI can help analyze the level of infection by this virus identifying the clusters and ‘hot spots’ and can successfully do the contact tracing of the individuals and also to monitor them. It can predict the future course of this disease and likely reappearance.

IV) Projection of cases and mortality

This technology can track and forecast the nature of the virus from the available data, social media and media platforms, about the risks of the infection and its likely spread. Further, it can predict the number of positive cases and death in any region. AI can help identify the most vulnerable regions, people and countries and take measures accordingly.

V) Development of drugs and vaccines:

AI is used for drug research by analyzing the available data on COVID-19. It is useful for drug delivery design and development. This technology is used in speeding up drug testing in real-time, where standard testing takes plenty of time and hence helps to accelerate this process significantly, which may not be possible by a human [6,7].

It can help to identify useful drugs for the treatment of COVID-19 patients. It has become a powerful tool for diagnostic test designs and vaccination development [[9], [10], [11]]. AI helps in developing vaccines and treatments at much of faster rate than usual and is also helpful for clinical trials during the development of the vaccine.

VI) Reducing the workload of healthcare workers

Due to a sudden and massive increase in the numbers of patients during COVID-19 pandemic, healthcare professionals have a very high workload. Here, AI is used to reduce the workload of healthcare workers [[12], [13], [14], [15], [16], [17]]. It helps in early diagnosis and providing treatment at an early stage using digital approaches and decision science, offers the best training to students and doctors regarding this new disease [18,19]. AI can impact future patient care and address more potential challenges which reduce the workload of the doctors.

VII) Prevention of the disease

With the help of real-time data analysis, AI can provide updated information which is helpful in the prevention of this disease. It can be used to predict the probable sites of infection, the influx of the virus, need for beds and healthcare professionals during this crisis.

AI is helpful for the future virus and diseases prevention, with the help of previous mentored data over data prevalent at different time. It identifies traits, causes and reasons for the spread of infection. In future, this will become an important technology to fight against the other epidemics and pandemics. It can provide a preventive measure and fight against many other diseases. In future, AI will play a vital role in providing more predictive and preventive healthcare.


[1]A. Haleem, M. Javaid, VaishyaEffects of COVID 19 pandemic in daily lifeCurr Med Res Pract (2020), 10.1016/j.cmrp.2020.03.011Google Scholar

[2]H.X. Bai, B. Hsieh, Z. Xiong, K. Halsey, J.W. Choi, T.M. Tran, I. Pan, L.B. Shi, D.C. Wang, J. Mei, X.L. JiangPerformance of radiologists in differentiating COVID-19 from viral pneumonia on chest CTRadiology (2020), 10.1148/radiol.2020200823Google Scholar

[3]Hu Z, Ge Q, Jin L, Xiong M. Artificial intelligence forecasting of COVID-19 in China. arXiv preprint arXiv:2002.07112. 2020 Feb 17.Google Scholar

[4]T. Ai, Z. Yang, H. Hou, C. Zhan, C. Chen, W. Lv, Q. Tao, Z. Sun, L. XiaCorrelation of chest CT and RT-PCR testing in coronavirus disease 2019 (COVID-19) in China: a report of 1014 casesRadiology (2020), 10.1148/radiol.2020200642Google Scholar

[5]H. Luo, Q.L. Tang, Y.X. Shang, S.B. Liang, M. Yang, N. Robinson, J.P. LiuCan Chinese medicine be used for prevention of coronavirus disease 2019 (COVID-19)? A review of historical classics, research evidence and current prevention programsChin J Integr Med (2020), 10.1007/s11655-020-3192-6Google Scholar

[6]A. Haleem, R. Vaishya, M. Javaid, I.H. KhanArtificial Intelligence (AI) applications in orthopaedics: an innovative technology to embraceJ Clin Orthop Trauma (2019), 10.1016/j.jcot.2019.06.012Google Scholar

[7]Biswas K, Sen P. Space-time dependence of coronavirus (COVID-19) outbreak. arXiv preprint arXiv:2003.03149. 2020 Mar 6.Google Scholar

[8]J. Stebbing, A. Phelan, I. Griffin, C. Tucker, O. Oechsle, D. Smith, P. RichardsonCOVID-19: combining antiviral and anti-inflammatory treatmentsLancet Infect Dis (2020 Feb 27)Google Scholar

[9]C. Sohrabi, Z. Alsafi, N. O’Neill, M. Khan, A. Kerwan, A. Al-Jabir, C. Iosifidis, R. AghaWorld Health Organization declares global emergency: a review of the 2019 novel coronavirus (COVID-19)Int J Surg (2020 Feb 26)Google Scholar

[10]S. Chen, J. Yang, W. Yang, C. Wang, T. BärnighausenCOVID-19 control in China during mass population movements at New YearLancet (2020), 10.1016/S0140-6736(20)30421-9Google Scholar

[11]S. Bobdey, S. RayGoing viral–COVID-19 impact assessment: a perspective beyond clinical practiceJ Mar Med Soc, 22 (1) (2020 Jan 1), p. 9CrossRefView Record in ScopusGoogle Scholar

[12]Gozes O, Frid-Adar M, Greenspan H, Browning PD, Zhang H, Ji W, Bernheim A, Siegel E. Rapid ai development cycle for the Coronavirus (COVID-19) pandemic: initial results for automated detection & patient monitoring using deep learning ct image analysis. arXiv preprint arXiv:2003.05037. 2020 Mar 10.Google Scholar

[13]B. Pirouz, S. ShaffieeHaghshenas, S. ShaffieeHaghshenas, P. PiroInvestigating a serious challenge in the sustainable development process: analysis of confirmed cases of COVID-19 (new type of coronavirus) through a binary classification using artificial intelligence and regression analysisSustainability, 12 (6) (2020 Jan), p. 2427CrossRefGoogle Scholar

[14]D.S. Ting, L. Carin, V. Dzau, T.Y. WongDigital technology and COVID-19Nat Med (2020 Mar 27), pp. 1-3View Record in ScopusGoogle Scholar

[15]K.H. Wan, S.S. Huang, A. Young, D.S. LamPrecautionary measures needed for ophthalmologists during pandemic of the coronavirus disease 2019 (COVID-19)Acta Ophthalmol (2020 Mar 29)Google Scholar

[16]L. Li, L. Qin, Z. Xu, Y. Yin, X. Wang, B. Kong, J. Bai, Y. Lu, Z. Fang, Q. Song, K. CaoArtificial intelligence distinguishes COVID-19 from community-acquired pneumonia on chest CTRadiology (2020 Mar 19), p. 200905View Record in ScopusGoogle Scholar

[17]A.W. Smeulders, A.M. Van GinnekenAn analysis of pathology knowledge and decision making for the development of artificial intelligence-based consulting systemsAnal Quant Cytol Histol, 11 (3) (1989 Jun 1), pp. 154-165View Record in ScopusGoogle Scholar

[18]R. Gupta, A. MisraContentious issues and evolving concepts in the clinical presentation and management of patients with COVID-19 infection with reference to use of therapeutic and other drugs used in Co-morbid diseases (Hypertension, diabetes etc.)Diabetes, Metab Syndrome: Clin Res Rev, 14 (3) (2020), pp. 251-254ArticleDownload PDFView Record in ScopusGoogle Scholar

[19]R. Gupta, A. Ghosh, A.K. Singh, A. MisraClinical considerations for patients with diabetes in times of COVID-19 epidemic. Diabetes & Metabolic SyndromeClin Res Rev, 14 (3) (2020), pp. 211-212ArticleDownload PDFView Record in ScopusGoogle Scholar

More information: Stephanie A. Harmon et al, Artificial intelligence for the detection of COVID-19 pneumonia on chest CT using multinational datasets, Nature Communications (2020). DOI: 10.1038/s41467-020-17971-2



Please enter your comment!
Please enter your name here

Questo sito usa Akismet per ridurre lo spam. Scopri come i tuoi dati vengono elaborati.