Artificial Intelligence, Data Science and Technology Solutions in the Fight Against Coronavirus

COVID-19 caused by the novel coronavirus is now known to spread primarily through people having “contact” with one another. This is precisely where technology can offer rapid solutions by reducing contact between people. In recent years, machine learning and artificial neural networks have been used in deep learning, an artificial intelligence approach increasingly being tested in medicine. Recent years have even seen research results showing AI applications making more accurate diagnoses than specialist doctors. Then the COVID-19 Pandemic became the focus of the global agenda, and the technology world began making major efforts to reduce its negative consequences. From China, where the outbreak began, and neighboring countries including Singapore, which set up mobile network infrastructure, we have seen examples of effective contact tracing. China is now tracking its citizens very strictly nationwide through both mobile networks and more than 2 million street cameras… In this article, we compile AI and biotechnology applications under 10 headings, from contact tracing and hospital robots to remote healthcare tools and algorithms facilitating new drug development.
1. Identifying, Tracking and Predicting Outbreaks
If we can track the virus and identify where it will spread in advance, we can fight it more effectively. Applications and algorithms developed for this purpose can identify and track outbreaks by analyzing news channels, social media platforms and government reports.
The Canada-based company BlueDot could be described as legendary in this field. Using more than a hundred data points, including flight tickets, weather and even animal numbers and species, it reported the outbreak days in advance and issued serious warnings about the necessary precautions. Its first warning came on December 31, 2019. This date matters because the World Health Organization did not issue its warning until January 9, 2020. On January 14, 2020, BlueDot also shared a travel warning called “Journal of Travel Medicine.” It listed the 20 cities receiving the most visitors from Wuhan. BlueDot’s researchers added that these cities could lead the spread of the virus.
In the near future, similar applications may warn us not only about the pandemic caused by the animal-origin COVID-19 virus, but also about human-caused problems such as climate change and environmental pollution.
AI can also be used for real-time tracking, as various examples in recent years have shown. In 2015, for example, a dynamic neural network model was used to track Zika as it spread from Brazil toward America. This algorithm could be retrained with current data and reused for the COVID-19 Pandemic, just as Carnegie Mellon University began using models predicting seasonal influenza spread after training them on new data. An important problem, however, is that the data obtained are sometimes insufficiently consistent. These systems are therefore used alongside epidemiological models.
These models are known as SIR. The SIR model is a way of calculating theoretically how many people will be affected during infectious disease outbreaks.
The name SIR:
- S: Susceptible – Likely to contract the disease
- I: Infected – Infected
- R: Recovered – Recovered
The name is formed from these initials. Data from this model are very important because governments base their decisions about precautions on them. The expanded SIR model used in China was shared with its code this past February so that other countries could benefit.
- Related topic: Seeking Solutions for COVID-19 Through Digital Footprint Tracking – How Does Each Country Monitor Its Citizens?
2. Artificial Neural Networks in COVID-19 Diagnosis
The virus’s genetic material is detected by RT-PCR molecular diagnostic tests, which are considered the main approach to coronavirus diagnosis but have limitations, such as obtaining samples, the time needed for analysis and accuracy problems. Alternative diagnostic methods are therefore being sought, including computed tomography (BT) (see Computed Tomography: The Most Effective Method in COVID-19 Diagnosis).
Certain distinctive radiological signs and imaging patterns specific to COVID-19 have been identified. It can take radiologists time to recognize these patterns, so AI algorithms are being applied to BT images to save time. Some studies have found that radiologists’ diagnostic times fell from about 30 minutes to as little as 5 minutes.
The faster the diagnosis, the more appropriate the intervention will be. The time saved may also help slow the outbreak’s spread and protect high-risk groups.
The Chinese e-commerce giant Alibaba says that its AI-assisted BT diagnostic system can diagnose patients with 96% accuracy.
3. Medical Equipment Support with Drones and Cleaning, Food and Logistics Tasks with Robots
The coronavirus’s ability to spread between people through the air has highlighted the importance of unmanned aerial vehicles and robots.
The Chinese company Terra Drone is helping provide logistical support between health centers in the Xinchang region using drones. Similar drones are also used to monitor quarantine rules and for thermal imaging.
Because of the outbreak, robots can be seen cleaning hospitals, caring for patients and distributing medicines and food. Pudu Technology, a food-focused company, has also installed its robots in more than 40 hospitals for these services (see the photograph below).

4. Developing New Medicines
Because producing a new medicine takes considerable time and money, scientists have focused on new treatment options using existing medicines. Richardson and colleagues, studying relationships between medicines and proteins through biomedical knowledge graphs, found that baricitinib, a medicine used for joint inflammation, might be effective against COVID-19.
Using a similar graph-based approach, Ge and colleagues applied natural language processing to the Pubmed database to examine relationships between human proteins, viral proteins and medicines. They found that the PARP1 inhibitor CVL218 might also be effective in treatment and began clinical trials.
Exscientia, a British start-up, became the first company to begin human trials of drug molecules designed with AI.
Using their algorithms, they designed a new medicine in only 12 months. With traditional methods, this process could take 4 to 5 years.
According to company executive Professor Andrew Hopkins, AI can be used in 3 areas of new drug discovery:
- Rapid production of antibodies and vaccines for Covid-19
- Screening existing medicines to identify appropriate treatment options
- Designing a new medicine for use in both the current and subsequent coronavirus outbreaks.
He adds that, at best, the process could take 18 to 24 months, depending on production planning and safety testing in people.
A group of researchers from South Korea and the United States has begun work on a commercial antiviral medicine for COVID-19, also using deep learning techniques. Like earlier studies, the team focused on drug interactions in the existing literature and announced that atazanavir, also used to treat HIV/AIDS, might be effective. Their study also identified gansiklovir, lopinavir and ritonavir, Prezcobix, darunavir and kobisistat as potentially active substances. They hope these findings and studies may help clinical trials being conducted in China and other countries.
Another group in China is using deep learning algorithms to understand better how the COVID-19 virus affects human cells. The researchers designed an advanced neural network model called DFCNN to identify and rank protein-ligand interactions with high accuracy. DFCNN identified potential medicines for the COVID-19 protease (the targeted part of the virus), and they published some potential drug groups.
Google-owned DeepMind used its AlphaFold system to predict coronavirus protein structures. These have not yet been experimentally validated, but DeepMind hopes the data may help explain the virus’s functions.

5. Producing Highly Protective Masks
The low level of protection offered by traditional masks against coronavirus has led researchers to seek new solutions. Companies such as Israel-based Sonovia, involved in producing antiviral masks using ultrasonic textile technologies, have begun making high-tech masks for healthcare workers and others. Another advantage is reusability: Sonovia’s masks retain their effectiveness against viruses even after 65 washes at 92 degrees.
- Related topic: Can Masks Be Disinfected and Reused?
6. Fighting Fake News and Misinformation
Google and Facebook are currently putting their technical teams to work intensively on identifying this misinformation and stopping its spread.
When you search Google for COVID-19, an information window opens in the browser’s side panel and offers links intended to inform you correctly.
YouTube also provides links in coronavirus-related content that take you directly to the World Health Organization’s (WHO) information page.
The spread of fake news not only misinforms users but also affects the accuracy of AI algorithms trained on these data to track and predict COVID-19 spread. Google Flu is one of the projects remembered as a failure, and manipulation of the data obtained is known to have contributed. Social media platforms therefore have a major role in this area.
7. Tracking and Identifying Infected People
Thermal cameras were being used to identify infected people, but their main disadvantage was that people still had to operate them. AI-assisted multisensor cameras have now begun to be used in communal places such as airports, hospitals and nurseries.
Following these developments, Google and Apple recently announced cooperation on COVID-19 tracking. Their project aims to use BLE (Bluetooth Low Energy) technology to examine whom smartphone users approached or had contact with for long enough to contract coronavirus.
One advantage is that users will not need to download an application from an app store. The method is intended to use the devices’ existing sensors. It aims to prioritize privacy by using no location or personal data.
The 3 main points emphasized in Apple and Google’s joint press release were “privacy, transparency and user consent.” They also announced that project information would be shared as open source after development.
Reaching certain testing levels is, of course, considered important for the project’s success.

Some countries, including China, Singapore, Israel, South Korea and Poland, have begun using tracking technology through applications installed on mobile phones to combat coronavirus. Similarly, Turkey, the United Kingdom, Italy and Germany continue working on their own digital initiatives.
Apple and Google aim to bring all these third-party applications together under one roof through their platform. This could allow the applications to work compatibly with one another and identify people within a shared coronavirus-response system even when they travel abroad.
If the application succeeds, countries may ease strict quarantine and isolation conditions or relax restrictions on travel and entry and exit.
The companies plan to announce their application interfaces (API) by mid-May. They have also published the necessary guidance for software developers to integrate other applications.
8. Answering Important Questions for Researchers
Kaggle, owned by Google and one of the world’s leading data scientist communities, launched a competition to answer COVID-19 researchers’ questions. Participants were given more than 50 thousand academic publications as source material and asked to apply natural language processing (NLP) methods to this dataset to seek answers to key questions. One of the competition’s main supporters is the White House Science and Technology Council.
Researchers from many fields worldwide are currently working very hard and publishing studies to combat coronavirus. The scientific literature is growing exponentially, and machine learning is believed capable of saving researchers time when reviewing these publications and answering some of their questions.
Apart from this project, competitions have also been launched on predicting COVID-19 cases and deaths and seeking solutions to problems frequently faced by doctors on the front line.
9. Remote Healthcare Services – Telehealth
At a time when discussion has begun about whether face-to-face examinations should be a last resort for meeting patients’ needs, shortages of intensive care specialists and doctors’ safety against COVID-19 have brought telehealth services back into discussion, and many companies and hospitals have begun collaborating.
Sheba Medical Center in Israel is one of the centers implementing innovations in this field. Patients diagnosed with COVID-19 are observed and treated in an isolation center 2 km from the hospital. This minimizes contact not only with patients quarantining at home but also with inpatients. Inpatients’ clinical data and movements can be monitored using AI-assisted sensors developed by EarlySense. Robots produced by InTouchHealth Robot are also used to conduct rounds.
Using examination kits produced by TytoCare, doctors can examine patients who need monitoring at home without contact, entirely remotely, at the desired place and time, and provide treatment plans and prescriptions.

10. 3-Dimensional Printing Technologies
During the pandemic, supply chains are experiencing serious difficulties providing medical equipment such as test kits and ventilator components. 3-dimensional printers are helping address this. In Ohio, for example, 250 printers normally used for dental prostheses have begun producing 100 thousand test kits daily. Many designers and engineers have prepared valves to increase the number of people able to use ventilators, sharing designs and user guides to enable worldwide use. A team at the University of Minnesota designed 6 critical components that make ventilators cheaper and quicker to produce and plans to share them in the coming days. Many car manufacturers in Europe and America have begun producing medical and protective equipment and continue sharing their experience and designs.
- ALL OUR CORONAVIRUS ARTICLES -
Sources
2. https://www.gpsworld.com/china-fights-coronavirus-with-delivery-drones/
3. https://www.bbc.com/news/technology-52246319
4. https://www.bbc.com/news/technology-51851292
5. https://www.kaggle.com/covid19
6. https://academic.oup.com/jtm/article/27/2/taaa008/5704418
7. https://bmcmedicine.biomedcentral.com/articles/10.1186/s12916-019-1389-3
8. https://www.medrxiv.org/content/10.1101/2020.02.29.20029421v1
9. https://www.tytocare.com/
10. https://www.technologyreview.com/2020/03/27/950240/3d-printing-coronavirus-covid-19-medical-supplies-devices/
11. https://www.preprints.org/manuscript/202002.0061/v1
12. https://towardsdatascience.com/covid-19-and-drug-development-b800d8917b50
Was this article helpful?
Your feedback helps us improve our content.





Reader comments
Share your comment
Your name and comment are published only after administrator approval. Your email address is never shown to visitors.
Be the first to comment.