Page 10 - AIH-1-2

Artificial Intelligence in Health AI in the battle against COVID-19

global and sociopolitical settings. It offers a critical ensure the selection of studies that provide robust and
evaluation of both successful and less successful AI relevant insights into the applications of AI during the
implementations. COVID-19 pandemic. These criteria serve as a safeguard
against methodological inconsistencies and form the
Section 11 looks forward to emerging technologies that
may influence the future role of AI in pandemic response. foundation for compiling evidence of high quality.
It provides policy recommendations to maximize the 4.2.1. Inclusion criteria
benefits of AI in this context.
The inclusion criteria encompass the following:
4. Methodology (i) Relevance to AI and COVID-19: Studies were included
if they explicitly addressed the deployment of AI
This comprehensive review employs a meticulous and technologies in the detection, diagnosis, treatment,
expansive literature search strategy designed to encompass or management of COVID-19, or in the analysis of
the full spectrum of AI applications in the context of the pandemic-related data.
COVID-19 pandemic. This strategy ensures the inclusion (ii) Peer-reviewed publications: Only peer-reviewed
of a diverse array of studies that provide a representative publications were considered, ensuring that all
cross-section of the current state of knowledge. included studies had undergone rigorous academic
4.1. Literature search strategy scrutiny and met the high standards of scientific
inquiry.
The development of our search criteria was a collaborative (iii) Empirical research studies: The review was confined
and iterative process, involving a consensus among a to empirical research studies that presented original
team of interdisciplinary researchers. A comprehensive data or analyses, providing concrete evidence of AI’s
search was conducted across multiple academic databases efficacy and utility in the pandemic context.
and search engines, including PubMed, Scopus, Web
of Science, and Google Scholar, to ensure a thorough 4.2.2. Exclusion criteria
survey of the existing literature. The search strategy was The review employed the exclusion criteria as follows:
augmented using Boolean operators, truncation, and (i) Non-English publications: Studies not published in
wildcard characters to maximize the retrieval of relevant English were excluded, given the linguistic capabilities
studies. of the review team and the need to ensure clarity and
The search was intentionally broadened to include consistency in the synthesis of findings.
studies from a multitude of disciplines, recognizing the (ii) Preprints and gray literature: Preprints and gray
inherently interdisciplinary nature of AI applications in literature were excluded to maintain a focus on
pandemic response. This approach facilitated the inclusion validated and peer-reviewed research, thereby
of research spanning the domains of healthcare, public upholding the review’s standard for evidence-based
health, computer science, and social sciences. conclusions.
The temporal scope of the search was defined to include 4.3. Data extraction and analysis
studies published from the start of the pandemic in late The data extraction and analysis phase are critical in the
2019 through to the present day. The search strategy was literature review process, where data is meticulously
periodically updated to incorporate the latest research gathered from selected studies and rigorously analyzed
findings, ensuring the review is up-to-date.
to form meaningful insights. This section elucidates the
A carefully curated list of keywords and topic headings methodical approach adopted for extracting and analyzing
was employed, encompassing terms such as “COVID-19,” data during the research process.
“SARS-CoV-2,” “artificial intelligence,” “AI,” “machine
learning,” “deep learning,” “neural network,” “pandemic,” 4.3.1. Data extraction protocol
“public health,” and “telemedicine,” among others. This Data were extracted from studies that met the inclusion
strategy was instrumental in unearthing studies that criteria, focusing on the application of AI in various aspects
specifically addressed the multifaceted applications of AI of the COVID-19 response globally. This information
in the pandemic milieu. included data on vaccine efficacy, treatment outcomes,
diagnostic accuracy, and predictive analytics. Standardized
4.2. Inclusion and exclusion criteria
data extraction forms were employed to ensure consistency
The integrity of this review is subject to a stringent set of and reliability across the data extraction process. These
inclusion and exclusion criteria, meticulously crafted to forms were designed to capture all relevant information,

Volume 1 Issue 2 (2024) 4 doi: 10.36922/aih.2401