Intelligent data and analytical solutions can help keep a vigilant eye on the emergence of pandemics and fortify healthcare supply chains, provided they follow diligent data processing methodology
The rising wave of the pandemic and the record number of cases has made it imperative for governments worldwide to identify innovative ways to track, detect, and diagnose COVID-19 cases and prepared for such a crisis in the future. Since the beginning of the outbreak, it has become challenging to track spikes in the coronavirus cases and predict their impact on the community.
The contagious virus resulted in uncertainty in every aspect of human life. And it has become soon evident that to tackle the gravity of the situation and be ready for such a future crisis, an extraordinary effort is required. In the last twelve to fourteen months, much research and analysis have been done on discovering the best ways to curb the coronavirus. However, even with stringent social distancing measures and accelerated vaccination programs worldwide, many countries are still struggling to prevent the alarming levels of virus spread and its severity.
In such a scenario, decision-makers are continually looking at technological interventions such as Artificial Intelligence (AI), Machine Learning (ML), and deep learning to make better predictions, act quickly, and minimize the impact of the pandemic. The AI-based epidemiological models have the potential to identify the possible hot spots of the disease swiftly and guide governments to accelerate health infrastructure and take proactive measures.
Today, millions of gigabytes of data and resources are available to track and identify pandemic-related patterns in a timely way. From infected patients' blood results to their age, sex, testing records, vaccination participation, treatment methodologies, availability of essential medicines and supplies, and outcome, a massive set of available data can help predict the future waves and analyze the overwhelming health care system. There are already some success stories, and many others are emerging. Data scientists are continuously evaluating the best ways to harness COVID-19 data using relevant algorithms and various simulations.
Early outbreak warning system
AI is no longer a marketing hyperbole. In recent years, deep learning and intelligent analytics have been effectively leveraged by companies and governments globally to accelerate production levels, analyzing disease patterns, and improve decision-making with a high degree of accuracy.
AI and ML have already proven their competence to detect an outbreak and even classify the locations that might need immediate attention. For instance, on December 31, BlueDot, a Canadian AI-based health monitoring platform, observed some strange pneumonia cases in Wuhan, China, using natural language processing and machine learning. Launched in 2014, the platform analyses over 100,000 news reports and studies about diseases in 65 languages every day. The platform analyses the data and consults with epidemiologists before sending warning signals.
The platform alarmed public health officials of several countries even nine days before the WHO released a warning statement around novel coronavirus. The same AI platform also identified India and Brazil as future epicenters months before the second wave hit in these countries.
Early outbreak warning systems are much needed to generate an immediate response and control the spread of COVID and impending epidemics. While these tools are still at an early stage of development, soon they can play a substantial role in identifying the emerging patterns of infectious diseases and inform global agencies and authorities to take proactive measures.
By scanning social media, different news articles, studies, and government data of various countries, these tools can analyze international data to forecast future events. While some of these AI systems can be fully automated, many others need constant human supervision to ensure that the correct data is being fed and recorded appropriately.
Tackling the outbreak
During the first wave, the focus of most of the countries was to detect the exposure, and hence contact tracing apps such as Aarogya Setu were launched to alert users about COVID-positive cases near them. There were few countries, though, which planned it much ahead to combat the crisis.
Taiwan, for instance, demonstrated its competence to leverage new-age technologies effectively to handle the outbreak and curb coronavirus transmissions early. It developed successful data intelligence templates to map the virus transmission and take proactive measures to contain it through AI and ML-led data analysis solutions and technology integration. By effectively collaborating with local technology research institutes and experts in automation and data science, the government of Taiwan unified its national health insurance database with the immigration and customs database.
The move helped the country tracked vulnerable and high-risk groups and ensured timely requisite measures to stop the community transmission. The AI-enabled algorithms helped the government improve the accuracy of COVID tests and enabled practitioners to analyze the impact the virus has made on the body organs of COVID patients.
With a growing focus on inoculation globally to curb the disease, countries like the US and the UK have leaped forward in using AI technology to meet the demand forecasting of medical essentials and supply chain management. Last year, UK's Medicines and Healthcare Regulatory Authority (MHRA) partnered with Genpact UK to deploy a machine learning software to capture the critical information regarding the adverse reactions of the COVID-19 vaccines. The software screens yellow card reports voluntarily submitted by patients and doctors, which entails unusual or side effects after taking COVID vaccination or administering medicines.
Similarly, IBM is helping the US government and hospitals to manage the supply chains of vaccinations through its Watson Health Analytics software. The AI-based software has been instrumental in predicting demand and ensuring vaccines are distributed fairly without any hiccups. In India, companies like Accenture and Microsoft collaborated with the Indian government to launch an AI-Chatbot, MyGov Saathi, to provide correct and latest health updates around COVID-19.
India's Defence Research and Development Organisation (DRDO) recently unveiled an AI-focused secure web-based COVID detection application software solution, ATMAN. The AI-based diagnostic tool helps rapid identification and analysis of lung condition of a patient by scanning chest X-rays and classify the images into Normal, COVID-19, and Pneumonia classes.
Learning from COVID-19
The COVID pandemic surge and the health crisis activated have changed everyone's perception of how healthcare services need to function. Across the world, the focus on integrating technology and innovative solutions to improve healthcare services has accelerated significantly. If we specifically talk about India, where the second wave of the pandemic resulted in an acute shortage of oxygen, medical supplies, and isolation beds.
Such a situation could have been averted or controlled better if the country had built a database of confirmed COVID leads. Then, through AI-based platforms, infected patients could get verified information, tracked, and asked to share about their complications to receive timely help from the government.
India's National Digital Health Mission program aims to create a digital health database of the country's citizens and develop necessary technological tools to address its need to improve its outreach to provide timely health care services.
Nevertheless, AI systems aren't unblemished. And while they offer tremendous potential to classify new illness types and transform the overall healthcare system, their effectiveness depends on data accuracy. The AI-systems algorithms make inferences from the data that is fed to them. Any gaps, stereotypes, or biases can present significant risks of giving misleading signals to enterprises, health officials, and the public. Then there are also concerns related to data privacy and security, preventing many users from sharing their accurate data for AI interpretations, hence needing to be addressed.
Building accurate healthcare databases and integrating them with the national identity number of individuals can serve as the foundation for developing robust AI-based healthcare tools. Such actions can minimize future surges of COVID, help improve the quality of healthcare systems and prepare us better respond to such outbreaks.