Page 11 - AIH-1-2

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

including study design, methodology, results, and neural networks and deep learning models has further
conclusions. refined the capabilities of AI, enabling the interpretation of
complex medical data with enhanced precision. 26,27
4.3.2. Analytical framework

The extracted data were synthesized to provide a 5.3. AI in genomics and drug discovery
comprehensive overview of the current state of AI in A notable milestone in the evolution of AI in healthcare
managing the COVID-19 pandemic. This synthesis is its application in genomics. 28-31 and drug discovery. 32,33
involved a qualitative assessment of the findings from the The completion of the Human Genome Project in the
included studies. Where applicable, a quantitative analysis early 2000s opened new avenues for AI applications in
was conducted to ascertain the effectiveness and impact of understanding genetic diseases and developing targeted
34
35
AI applications. This process involved statistical techniques therapies. AI-driven platforms such as AtomNet have
to combine data from multiple studies, providing a more since been utilized to identify potential drug candidates,
robust understanding of AI’s role in the pandemic. significantly reducing the time and cost associated with
traditional drug discovery processes.
5. Evolution of AI in healthcare
5.4. AI-enabled medical devices and wearables
The evolution of AI in health-care represents a significant
shift in medical practice and research. From early rule- The emergence of AI-enabled medical devices and wearables
based expert systems to deep learning models that leverage has significantly benefited patient monitoring and health
vast healthcare data and advanced analytics techniques, AI management. Devices such as smartwatches and fitness
has found its application in multifaceted areas of healthcare trackers, equipped with biomedical sensors and AI
and medicine. 18,19 This section delineates some of the algorithms, can now provide real-time insights into an
early developments of AI in medical diagnosis, genomics, individual’s health status, detecting anomalies that may
36
drug discovery, medical devices, and wearables. These require medical attention. These advancements have
advancements and research have laid a foundation on not only enhanced preventive healthcare measures but
which current technologies have been honed and adapted have also empowered individuals to take an active role in
in the fight against the COVID-19 pandemic. managing their health.
5.1. Rule-based expert systems 5.5. The role of AI in pandemic response
The inception of AI in health-care can be traced back to There were no major pandemics before the COVID-19
the early experiments with rule-based expert systems. pandemic where AI was used extensively or prominently
One such expert system is MYCIN from the 1970s, in the response. This is primarily because the development
designed to diagnose bacterial infections and recommend and widespread adoption of advanced AI technologies,
antibiotics. 20,21 Another significant system was the particularly in healthcare, coincided with or followed the
Internist-I (later developed into CADUCEUS), created in COVID-19 pandemic. Previous health crises, such as the
the late 1970s. This system focused on internal medicine H1N1 influenza pandemic in 2009 or the Ebola outbreak
22
and could diagnose complex cases by comparing patient in 2014 – 2016, occurred before AI had reached its current
data against a large database of disease profiles. Internist- level of sophistication and integration in health-care
I’s comprehensive approach to diagnosis showcased systems. During these earlier health crises, the use of AI
the potential of AI systems to handle a wide range of was either very limited or not a significant component of
medical knowledge. These pioneering efforts established the public health response.
the early relationship between computational algorithms However, it is noteworthy that before COVID-19,
and medical expertise, paving the way for advanced research efforts were made to explore the potential use
AI applications in modern healthcare, where machine of technology and AI in disease outbreaks. Predictive
37
learning and data-driven approaches are now integral. modeling and data-driven techniques have been studied
to predict infectious disease epidemics. 38,39 Other studies
5.2. Integration of machine learning
demonstrated the use of machine learning analysis of
The integration of machine learning algorithms marked a social media and media sources for tracking public health
significant evolution in AI’s application within healthcare. trends and understanding public awareness during health
The shift from rule-based systems to data-driven approaches crises. 40,41 These studies collectively illustrate the evolving
allowed for the analysis of large datasets, leading to more role of AI, big data, and machine learning in monitoring
accurate diagnostic tools, personalized treatment plans, and predicting disease outbreaks, offering valuable insights
and predictive analytics. 23-25 Notably, the development of for pandemic preparedness and response.

Volume 1 Issue 2 (2024) 5 doi: 10.36922/aih.2401