Covid-19: How Tech is Helping

I don’t have to say anything we all know, this is the unpredicted time since the first report of covid-19 in Wuhan, China was reported and as time passed it has spread all over the globe. As China grappled with the crisis, the world watched. It seemed distant to most but then a wave of infections began to appear in our city, state and country. Governments, health officials and public scrambled to make sense of situation and drastic changes like lockdown were made. For many the ceasing of everyday life has become a surreal shock. It’s a difficult time for most as confusion and uncertainty prevails. There certainly has been a chaos but also a hope.

In this blog we are going to see how people are coming together in using technology to fight this pandemic with the help of Artificial Intelligence to sterilizing robots to rapid development of new technologies being reappropriated in ways we have never seen before. Governments are developing and modifying policies old and new to promote the rapid development of technologies that can help eliminate the Coronavirus. Usually emerging technologies are held back by infrastructure, financing and bureaucratic constraints but with the challenge of Covid-19 we can put these new technologies to the test.

Identifying who is most at risk from COVID-19 with the help of Machine Learning:

Machine learning has proven to be valuable in predicting risks in many spheres. With medical risk specifically, machine learning is potentially interesting in three key ways.

Infection risk: What is the risk of a specific individual or group getting COVID-19?

Severity risk: What is the risk of a specific individual or group developing severe COVID-19 symptoms or complications that would require hospitalization or intensive care?

Outcome risk: What is the risk that a specific treatment will be ineffective for a certain individual or group, and how likely are they to die?

Machine learning can potentially help predict all three risks. Although it’s still too early for much COVID-19-specific machine learning research to have been conducted and published, early experiments are promising. Furthermore, we can look at how machine learning is used in related areas and imagine how it could help with risk prediction for COVID-19.

Robots- Sterilizers and Delivers:

The idea of robots taking up jobs previously done by humans was seen as a threat even a few months ago. But now everyone believes that robots could perform some of the “dull, dirty and dangerous” jobs associated with combating the COVID-19 pandemic, in the areas where humans cannot step in. Obviously robots aren’t susceptible to the virus so they are being deployed to carry out many tasks such as cleaning and sterilizing and also delivering food and medicine. Robots can also be used for clinical care such as de-contamination, delivery and handling of contaminated waste as well as monitoring compliance with voluntary quarantines. This is all in effort to reduce human to human contact.

Hong Kong’s Mass Transit Railways is now employing vaporized Hydrogen peroxide robots to disinfect their trains. The Transit Railway services millions of passengers each day and this has become a breeding ground for the virus. These newly purchased deep cleaning robots will be able to reach places which will be difficult to get by hands. They are also going to be deployed where covid-19 patients have been making it safe for human entry.

In India, the Sawai Man Singh Government Hospital in Jaipur, Rajasthan is conducting a series of trials on a humanoid robot to check if it can be pressed into service for delivering medicines, and food to the COVID-19 patients admitted there. This could potentially reduce the chances of the hospital staff contracting the infection.

Another company, Asimov Robotics, a Kerala-based startup has developed a three-wheeled robot that it says can be used to assist patients in isolation wards. This will include helping with things like food and medication, something that nurses and doctors have been doing so far, putting them at larger risk of contracting the virus.

UVD Robots: 

The UVD Robot is used as part of the regular cleaning cycle, and aims at preventing and reducing the spread of infectious diseases, virus, bacteria, and other types of harmful organic microorganisms in the environment by denaturation of their DNA-structure. The robot is safe, reliable and eliminates human error. Many hospitals are using an ultraviolet (UV) light robot to disinfect the facility curbing the potential spread to other humans. The hospital is using UV light instead of hydrogen peroxide, because it cuts disinfecting time down from hours to ten to fifteen minutes and also it avoids human contact with the infected environment. For disease prevention, robot-controlled non-contact UV surface disinfection has already been used because COVID-19 spreads not only from person to person via close contact respiratory droplet transfer but also via contaminated surfaces.  

In China, Disinfection Robot UVD was in high demand since the outbreak of the pandemic. A large number of hospitals in the country have been ordering the robot which is manufactured by Denmark’s Blue Ocean Robotics. These robots have played a key role in controlling the virus in Wuhan, the epicenter of the virus.

Figuring out how to attack the virus

Epitopes are clusters of amino acids found on the outside of a virus. Antibodies bind to epitopes, which is how our immune system recognizes and eliminates the virus. So finding and classifying epitopes is essential in determining which part of a molecule to target when we develop the vaccines. Compared to traditional vaccines, which contain inactivated pathogens, epitope-based vaccines are safer – they prevent disease without the risk of potentially deadly side effects. Locating the correct epitope can be a time-consuming, expensive process.

With a new pandemic, such as COVID-19, locating epitopes faster speeds up the process of developing effective vaccines. This is where machine learning can help. Support vector machines (SVM), hidden Markov Models, and artificial neural networks (specifically deep learning) have all proven to be faster and more accurate at identifying epitopes than human researchers are.

3D Printing to develop Ventilators: 

This is the great example of what can happen when smart and technically oriented people come together in a time of need. Ventilators have become essential in the battle against covid-19 but health systems around the world are facing shortages. To address this problem, engineers around the world have set up network platforms using technology platforms such as Telegrams and Facebook. Here they share information about open source designs for manufacturing ventilators with 3d printers. Anyone with a 3d printer can collaborate by printing the necessary respirator components. The goal is to make them available to health care services.

To add to this many companies such as Tata, Tesla, Ford, General Motors are all pitching in to make ventilators. In India many technical and research universities such as IIT Hyderabad, IIT Kanpur and others are making low cost ventilators to help India win the battle against Covid-19.

Role of Artificial Intelligence: 

Diagnostic AI:

Immediate diagnosis means that response measures such as quarantine can be employed quickly to curb further spread of the infection. An impediment to rapid diagnosis is the relative shortage of clinical expertise required to interpret diagnostic results due to the volume of cases. AI has improved diagnostic time in the COVID-19 crisis through technology such as that developed by LinkingMed, a Beijing-based oncology data platform and medical data analysis company. Pneumonia, a common complication of COVID-19 infection, can now be diagnosed from analysis of a CT scan in less than sixty seconds with accuracy as high as 92% and a recall rate of 97% on test data sets. This was made possible by an open-source AI model that analyzed CT images and not only identified lesions but also quantified in terms of number, volume and proportion. This platform, novel in China, was powered by Paddle Paddle, Baidu’s open-source deep learning platform.

Disease Surveillance AI:

The detection of outbreak and the issuance of public warnings are critical during pandemics. BlueDot a Canadian startup developed an AI which analyzes news and government reports as well as social media in order to track infectious diseases. BlueDot had already issued a warning before the World Health Organization (WHO). Unfortunately not many people listened to these warnings. BlueDot published the first scientific paper on COVID-19, accurately predicting its global spread using our proprietary models.

With an infectious disease like COVID-19, surveillance is crucial. Human activity -especially migration- has been responsible for the spread of the virus around the world. In the near and distant future, technology like this may be used to predict zoonotic infection risk to humans considering variables such as climate change and human activity. The combined analysis of personal, clinical, travel and social data including family history and lifestyle habits obtained from sources like social media would enable more accurate and precise predictions of individual risk profiles and healthcare results. While concerns may exist about the potential infringement to civil liberties of individuals, policy regulations that other AI applications have faced will ensure that this technology is used responsibly.

Virtual Health Assistans(Chatbots): 

The number of COVID-19 cases has shown that healthcare systems and response measures can be overwhelmed. Canada-based Stallion.AI has leveraged its natural language processing capabilities to build a multi-lingual virtual healthcare agent that can answer questions related to COVID-19, provide reliable information and clear guidelines, recommend protection measures, check and monitor symptoms, and advise individuals whether they need hospital screening or self-isolation at their homes.

Facial Recognition and Fever Detector AI 

Thermal cameras have been used for some time now for detecting people with fever. The drawback to the technology is the need for a human operator. Now, however, cameras possessing AI-based multisensory technology have been deployed in airports, hospitals, nursing homes, etc. The technology automatically detects individuals with fever and tracks their movements, recognize their faces, and detect whether the person is wearing a face mask.

Another useful resource is ResApp. ResApp Health is developing digital healthcare solutions to assist doctors and empower patients to diagnose and manage respiratory disease. It uses the machine learning algorithms to analyze a patient’s cough sound to treat diseases. They are hoping to use this system in the battle against Covid-19.

Predicting protein folding: 

We know that a protein’s structure is linked to its function – and once this structure is understood, we can guess its role in the cell, and scientists can develop drugs that work with the protein’s unique shape. But defining a protein’s 3D structure is no easy task – the range of possible structures for a single protein is astronomical: a protein composed of 100 amino acids has 3100 possible conformations.

unfolded vs folded

And there are over one billion known protein sequences, but we have only been able to identify the structures of less than 0.1% of them. Using artificial neural networks, research groups have successfully built models that can predict protein structures, finally making it feasible to identify protein structures using computational methods.

Role of Active learning (AL) 

As compared to the passive learning (traditional machine learning classifiers), active learning is used to a learning problem, where the learner has some role in determining on what data it will be trained [13]. When it is an emergency (COVID-19) [6], it requires a special attention so that data analysis and decision-making can be made consistently without waiting several days, months, and years for data collection. Again, exploiting real-time data (on-the-fly) is the must since one cannot wait for years to train machine and learn from them nor manual annotation/analysis is possible. This means that instead of having a conventional set of train, validation, and test set, we need AI-driven tools that can learn over time without having complete knowledge about the data, which we call Active Learning (AL). In other words, AL mechanism helps self-learn i.e., Incremental Learning (IL) over time in the presence of experts (if required) [18]. The ILs aim is to iteratively help learn model to adapt to new data without forgetting its existing limited knowledge. Figure ​Figure33 provides a schematic diagram of an AL mechanism, where different data types are used. While learning, the changes in data over time can be assessed with the help of Anomaly Detection (AD) techniques. In AL-based tool, AD helps find/identify rare items, events or observations that bring suspicions by differing significantly from the majority of the data or with respect to a set of normal data for that particular event.

For time-series data, a schema of Active Learning (AL) model is provided. For better understanding, AL (in dotted red circle) is used with Deep Learning (DL) for all possible data types. In AL, expert’s feedback is used in parallel with the decisions from each data type. Since DL are data dependent, separate DLs are used for different data type. The final decision is made based on multitudinal and multimodal data 

 UiPath’s New Technology: 

To support the healthcare crisis, UiPath, an enterprise Robotic Process Automation (RPA) software company, is offering healthcare organizations free RPA software to accelerate critical processes and free up strapped employees so they can more rapidly respond to issues arising as a result of the COVID-19 pandemic. The Mater Hospital in Dublin is using UiPath’s attended robots to process COVID-19 testing kits in a fraction of the time. The hospital not only receives patients’ results in near real time, but significantly reduces the administrative strain placed upon its Infection Prevention and Control Department. By giving a robot to every nurse, the department saves three hours per day so medical personnel can spend more time taking care of patients rather than completing paperwork.

Big Data and Tracking: A battle of Safety vs Freedom:

This technology will track the phone of the infected allowing for the government to track their GPS location. Then the government organization would exactly know who they have come into contact with and where they have been.

Israel is planning to repurpose their anti-terrorism technology in order to track the spread of the virus, although this is pending cabinet approval. United States of America is also engaging with tech giants to come up with similar tech in their country.

In Asia, countries like Vietnam is tracking locals and foreigners through mobile apps while Thai immigration authorities are using location data to track those are arriving in country. 

In India similar situation is rising after the India's new contact tracing app named as Aarogya Setu, created by the government’s National Informatics Centre may serve as a lesson in those privacy pitfalls: it could reveal the location of Covid-19 patients not only to government authorities but to any hacker clever enough to exploit its flaws. One specific feature of the app, designed to let users check if there are infected people nearby, instead allows users to spoof their GPS location and learn how many people reported themselves as infected within any 500-meter radius. In areas that have relatively sparse reports of infections, hackers could even use a so-called triangulation attack to confirm the diagnosis of someone they suspect to be positive. (If you want to know more about this app, make sure to write me in the comment section, I will cover the whole idea in the next blog.)

Thought this specific strategy of tracking has its ups and downs but it will surely help to slow down the rate of virus with more accurate and precise data you could have otherwise.

Video Calling:

As most of the people in the world are in the lockdown stage video calls are keeping people together. They have become an essential tool for dealing with confinement. One of the prominent application that boomed in this lockdown period is ZOOM, which approximately received 600,000 Downloads in one day at the beginning of the pandemic. Despite reports of privacy concerns the application is used to organize team meetings or helps for online school classes. But out of this use Zoom bombing is gaining attraction. Zoom bombing is something where are person barges into the office meeting or school classes.

Conclusion:

While this outbreak is something to take seriously humanity has persevered through the difficult times. The ways some countries are working together to stop this pandemic is uplifting and emotional to see. Scientists and Researchers have no longer geographical boundaries and the world is starting to put this technology to make a good use. So in finishing the blog please remember we must look out for each other, we must pull out each other through this. Because there’s one only thing that is more contagious than this virus it is love and compassion. 

Peace.

- Saurabh M

* DM for credits or removal request for images( no copyright intended ) 🔗 All rights and credits reserved to the respective owner.