COVID-19 is arguably the single biggest pandemic in recent human history. McCall (2020) attempts to argue the case for applying Artificial Intelligence (AI) to various domains surrounding the identification and prediction of the COVID-19 outbreak. The researcher further outlines several contemporary AI applications and sources of information relative to the previous comparable outbreak. This essay summarily reviews McCall’s perspective on AI application in COVID-19 management, relative to other AI implementation arguments in healthcare.
The current COVID-19 outbreak can be compared extensively to the Severe Acute Respiratory Syndrome (SARS) of 2003. However, the magnitude of infections and mortality rate of COVID-19 severely eclipse that of the SARS despite the epicenter of both outbreaks being China. McCall (2020) contrasts the two diseases and how they spread across continents, use similar mechanisms to infect the cell, and affect animals and humans. However, the researcher also appreciates the significant tactical development in the 17 years between the two diseases in the form of Artificial Intelligence (AI).
AI is causing a significant paradigm shift in healthcare. This particular assertion is prevalent in the extant literature, including Davenport and Kalakota (2019), Jiang et al. (2017), Maddox et al. (2019), Reddy et al. (2019), and is reiterated by McCall (2020). The researcher outlines that there may be value in applying AI to the current pandemic, especially in mapping the prevalence and predicting the spread of the epidemic to other locations. Contextually, this application of AI is seen in Blue Dot, a Canadian company credited as the first organization to break the news of the pandemic in late December. But the overall question remains as to whether the capability of AI is currently at a point to deliver compelling insight in a timely, widescale fashion.
There are several primary factors necessary for an effective public health intervention to a new viral outbreak. These include comprehension of the natural history of infection, risk populations and the causative organism, and the development of preventative and control measures from epidemiological modeling (McCall, 2020). Individuals primarily collect this information at outbreak sites that are virtually connected to WHO and represent a primary source of information for COVID-19. This data can reasonably be used to prime AI to read the evidence and link it effectively to outbreaks. Further, through the review of newsfeeds, social media, and airline ticketing systems, health professionals can identify outbreaks, and areas that need further exploration. However, this data is contingent on health systems with good contact-tracing and patient isolation protocols.
AI can also be used to make a significant contribution to the current pandemic in the prediction of how COVID-19 is affected by seasonality. Based on the premise of historical coronaviruses behavior, such an application could significantly help stabilize financial markets by reassuring the gradual diminishing of the epidemic. However, the efficacy of AI application to COVID-19 is outlined as ‘garbage in garbage out’ to indicate that the quality of data is significantly correlated with the insight gleaned.
China is not only the epicenter of the COVID-19 outbreak but a pioneer and supporter of AI application in helping to manage the epidemic. For instance, Infervision, an AI company located out of Beijing, China, developed a proprietary algorithm to detect COVID-19 on CT scans of the human lung distinctly and distinguish it from other respiratory infections. The application of this AI technology expedites COVID-19 diagnosis and monitoring, which further alleviates the need for governments implementing business and country lockdowns. Finally, the increased utilization of AI in reading scans allows it to learn and improve its accuracy significantly.
The death of Li Wenliang, a medical doctor and whistleblower on the COVID-19 epidemic, was indicative of the need to protect clinicians and healthcare professionals on the frontline. Hospital-associated transmission rates, for instance, from one human to another in Wuhan University’s hospital accounted for 41% for all cases, and a thousand hospital staff were confirmed infected as well (McCall, 2020). The application of AI could ideally help protect clinicians, hospital staff, and healthcare professionals.
While a doctor can manually read a CT scan for up to 15 minutes, Infervision, the Beijing-based diagnostic AI, can read it in 10 seconds. The AI detects lesions stemming from coronavirus pneumonia and provides measurements and comparative changes with other lesions. This provides sufficient quantitative data for doctors to make a prompt decision. AI-based CT imaging could serve as a stopgap measure for doctors whenever urgent diagnosis and judgment is required. High-risk cases can be promptly identified and removed from general areas to infect patients and hospital staff.
Scholarly discourse does concede that Artificial Intelligence is still a highly novel application within healthcare, and may take some time for extensive integration. It is yet relatively early to accurately determine the capability and extent, to which AI application will impact COVID-19. However, as the mortalities and infections rise, then so does the supply of research data. Overall, AI is relatively significant to this outbreak now, and perhaps even more so in the future.
References
Davenport, T., & Kalakota, R. (2019). The potential for artificial intelligence in healthcare. Future Healthcare Journal, 6(2), 94–98.
Jiang, F., Jiang, Y., Zhi, H., Dong, Y., Li, H., Ma, S., Wang, Y., Dong, Q., Shen, H., & Wang, Y. (2017). Artificial intelligence in healthcare: Past, present and future. Stroke and Vascular Neurology, 2(4), 230–243.Maddox, T. M., Rumsfeld, J. S., & Payne, P. R. O. (2019). Questions for artificial intelligence in health care. JAMA, 321(1), 31–32.
McCall, B. (2020). COVID-19 and artificial intelligence: Protecting healthcare workers and curbing the spread. The Lancet Digital Health, 2(4), e166-e167. Web.
Reddy, S., Fox, J., & Purohit, M. P. (2019). Artificial intelligence-enabled healthcare delivery. Journal of the Royal Society of Medicine, 112(1), 22–28.