In the midst of rapid environmental disruption caused by globalization and climate changes, the influence and consequence of infectious disease outbreaks have been accelerating unprecedently to become a major threat to individual health and social well-being.
In particular, as evidenced by recent Coronavirus Disease (COVID-19) pandemic, careful monitoring and timely intervention on infectious disease outbreak is essential to minimize morbidity and mortality. Thus, a dire need to develop a system to deliver rapid triage, diagnosis, and treatment arises, to overcome current global crisis.
Since various methodologies using artificial intelligence (AI) have proven efficient and effective in industries including healthcare, many attempts have been made to apply AI technology to better understand current COVID-19 crisis, and develop diagnostic and interventional tools. At the same time, we have observed the emergence of many technical pitfalls and challenges such as reliability of the data, trustworthiness of applications, and privacy protection.
In this report, we summarized the cases of infectious disease outbreak responses using AI-driven technology and trends of related standardization efforts. We also investigated essential elements needed to construct the AI-driven infectious disease response systems at the national and global level.
Infectious disease outbreak, COVID-19, Coronavirus, SARS-CoV-2, infectious disease, artificial intelligence, standardization, computer-aided diagnosis, clinical decision support system, machine learning, prediction
DOI:10.22648/ETRI.2020.B.000005
(Conceptual Stage) Research on AI-based tools for infectious diseases outbreak is in its infancy. However, we observe a rapidly growing body of literature with international collaboration on research and development.
(Necessity of AI-based infectious disease response system) This collaboration is based on the consensus that the infectious disease outbreak is one of the great challenges that human civilization encounters now.
(Artificial Intelligence - an excellent tool for infection diseases response system) With the enormous power to recognize characteristic patterns from complex data, the AI-based tools provided excellent performances throughout the entire process of infectious diseases response system in current COVID-19 crisis, including early prediction of the disaese outbreak, development of diagnostic technology, tracking of contacts, and the development of new therapeutic agents. In addition, Post-COVID-19 pandemic, we anticipate the contribution of AI-based technology will continue in advancing the national disease monitoring system and improving public health at an individual or community level.
(Expanding Horizon for AI-based Technologies and Applications) AI-based technologies including time-series analysis, natural language processing, and speech recognition are rapidly converging and developing. Continuous investment and robust research and development plans will emerge in areas with the highest level of commercialization and market potential, which includes clinical decision support system, remote patient monitoring, prognosis prediction, automatized response to disaster, contact tracing, and drug discovery research.
(Constraints and limitations) We observe the regulation and restrictions on the use of medical data for AI-based researches. May of the prediction model performances are not very reliable, thus cannot be used in the medical field. Only few models have reached at technical maturity, and a lot of cross-validation or external validation results have not been published. It is essential to recognize these limitations before developing AI-based tools.
(Privacy and Ethics) In order to achieve the desired level of quality in all aspects through development, not only securing and processing high-quality data, but also resolving privacy protection and ethical issues are important.
(Multi-disciplinary Cooperation) Multidisciplinary collaborative research efforts involve a team consisted with clinical experts including infectious disease specialists, artificial intelligence experts, as well as biology domain experts.
(Innovative Approaches) Infectious disease outbreak is not confined to a single country. International cooperation needs to be open and multi-directional to achieve a timely response to a rapidly evolving and spreading disease. The importance of open data, open source, and open science cannot be emphasized more, as these open cooperation and innovative efforts will further accelerate the evolution of the AI-based models and foster the growth of related industries.
(Overcoming Disparities Among Countries) A joint effort through international collaboration is required to tackle multiple agendas for consensus, such as reliability, standardization, interoperability, quality of the data as well as standard for interpretation. Country-specific rules and regulations should be discussed beforehand.
(Establishing an International Rapid Response System) A system that can quickly share and utilize high-quality data is needed, and an open global repository for sharing anonymous medical images and clinical data should be discussed. Standardized operational procedures and data protection systems to securely transfer/distribute/use data should also be established.
(Application of the Standardization System) In order to better respond to new infectious diseases in the future, a standardized interoperable system to plug AI-based tools and allow analysis and interpretation need to be agreed upon. One of these early efforts could entail the modification and application of the infection control system developed by the Centers for Disease Control and Prevention in Republic of Korea, which has been also partially related to the AI-based research projects and operational protocols.