Artificial intelligence is swiftly marching into an array of domains, and Medicine is no exception. With insanely tremendous advances in big data and computation, AI plays an instrumental role in revolutionizing the epidemiological investigation of infectious disease outbreaks.
Humanity has, time and again, faced infectious disease outbreaks. During the early 19th century, while looking for patterns of disease occurrence with a hope to prevent them, the researchers sowed the seeds of Epidemiology. And it had a paramount impact on the way medical science has evolved.
Florence Nightingale was one of the nurses sent by the British military to look after injured soldiers during the Crimean War in 1854, from her meticulous record keeping, looked on a vast majority of casualties because of infectious diseases. Little did she know, her data would later become an important epidemiological tool and lead to reforms in the British military medical care system to prevent infectious diseases caused by poor sanitation and inaccessible hospital facilities?
Thus, even if diseases like Tuberculosis have been with us for eons, in recent times we have witnessed the emergence of novel communicable infections with the potential to wreak global havoc, the ongoing COVID-19 being one of them. So, Artificial intelligence is touted as a promising tool being brought into the battle against such outbreaks.
Artificial Intelligence
Artificial intelligence (AI) refers to the capabilities of a computer or other machines to exhibit human-like intelligence to perform functions similar to what a human mind can perform. Learning from experience, understanding and responding to languages, making correct decisions, solving problems, recognizing objects, etc. collectively depict the potential of Artificial intelligence.
Artificial Intelligence Joining The Dots In The Field Of Infectious Disease Outbreaks
The current technological advances in Big Data & Analytics have paved the way for Artificial intelligence in the medical field as well. Artificial intelligence is especially being effectively deployed for gaining real-time insights into disease transmission and forecasting fresh outbreaks. Let’s see how?
AI-Enhanced Early Detection Of An Outbreak
Early warning of an outbreak ensures “safety nets” by identifying signs of unseen outbreaks and effective resource deployment to contain potential pandemics. Subsets of Artificial intelligence applications viz., Machine Learning (ML), Natural language processing (NLP), and Deep Learning are used to explore indicators of outbreaks from the community.
An AI-enabled machine, when fed with enormous data from the physical and cyber world, identifies trends and patterns in the data and senses red flags in them. Social media, news reports, and online data have served as the goldmine to spot localized outbreak even before it escalates to a greater magnitude.
The Canadian setup, Blue Dot successfully used AI and Machine Learning algorithms to locate a mysterious virus, now called COVID-19 emerging in Wuhan, China by the end of December 2019 and subsequently warned the travelers even before the WHO sent out the word to all the countries.
Also, Big Data analysis of medical records, satellite imaging showing crowding around the hospital, has helped to detect the early signs of outbreaks. Google trends can fuel data for dynamic forecasting models. Sentiment analysis using NLP in social media helps understand the emotions of the population, thus guides the government towards public awareness
In a nutshell, AI-based data analysis techniques not only help detect and curb outbreaks in their nascent stage but also provide the medical teams and other relevant authorities, buffer time to act towards prevention and management.
AI-Driven Predictive Modeling For Prediction Of The Spread
Big Data and AI-driven predictive modeling techniques are mathematical, statistical, or dynamic models that predict the extent of transmission in the population.
Unlike traditional models, it has added advantages viz., flexibility, recalibration based on trends, adaptive learning, the possibility of improving depending on the latest development in research, and evaluation of the efficiency of the intervention methods applied to control the spread like social distancing.
Susceptible-Exposed-Infectious-Recovered (SEIR) modeling is being used for estimating COVID-19 spread and other parameters like under-reporting of the cases and testing accuracy. Moreover, Backward Propagation Neural Network (BPNN) proved to be the most accurate in predictive modeling Zika virus transmission.
AI-Enhanced Case Tracking
No other epidemic has reinforced the gravity of case tracking the way COVID-19 did. Undoubtedly, early case tracking and quarantine helps prevent exposure to the rest of the population.
While case tracking, Artificial Intelligence tools can be deployed on Big Data from health insurance, customs, and immigration to track, and stratify infected individuals for further disease management protocols and quarantines.
Digital phenotyping is one such novel concept that collects data on smartphones and produces individual profiles. For example, “Arogya Setu”, the mobile app in India tracks and signals its users to exposure of COVID-19 patients in their immediate vicinity.
Also, research suggests Deep Learning algorithms as a promising way to track patterns of COVID-19 infection in CT and MRI imaging results. Their use in radiologic image processing can reduce false positive and negative errors and prove to be a powerful diagnostic method.
Summing it up!
Artificial intelligence offers a more exciting array of tools for rapid analysis and pattern identification of extensive data while studying the disease process compared to traditional tools. Their flexibility and self-learning ability make them highly versatile epidemiological weapons at any stage of infectious disease outbreaks.
