This post was originally published by Shadeeb Hossain at Towards Data Science
We cannot control the spread without AI technology
The rate of progress in Artificial Intelligence has been exponential in the past decade. The pandemic has made us more reliant on technology as we practice social distancing and work-from-home options. Artificial Intelligence and machine learning are providing options to utilize data from various sectors and make informed decisions. One such application is how AI is used to help us fight COVID-19.
Artificial Intelligence is actively used in identifying high risk patients at an earlier stage and therefore helps to control the spread of the infection in real-time. This becomes particularly important at this time of crisis because real time monitoring is the best option for people to self-isolate and mitigate the spread of the virus.
Clevy is a French startup that uses augmented assistance to help diagnose COVID-19 symptons. This can allow diagnosis without having to leave your home premises. It is a both time and cost-effective solution. Moreover, it reduces the time load and assist the healthcare community to cater to a larger population during this time of crisis.
Mayoclinic is another platform that is allowing self-assessment of COVID-19 symptoms and suggest ways to protect oneself during this pandemic.
AI can also be used to develop new diagnosis and management systems from COVID-19 cases across the globe. AI and Machine Learning made it easier to screen through a series of relevant previous and current studies related to COVID-19 and other relevant outbreaks.Machine Learning also is actively used to scale up customer communication ( through rob-chat options) and even help organize and collect relevant data related to COVID-19 research.
MIT-IBM Watson AI lab are actively involved in a series of AI related projects to mitigate the spread of COVID-19 pandemic. Almost all of those project use AI and ML platform.
One of their projects include the early detection of sepsis in COVID-19 patients. Patients with sepsis are less likely to survive when infected with COVID-19. The researchers plan on using ML to analyze the images of white blood cells (WBC) of COVID patients for an activated immune response against sepsis. Sepsis is a life threatening immune response to an infection. Only half of the patients who get sick with sepsis and have COVID-19 survive ( it is a deadly combination). This early diagnosis will allow time for physicians to take necessary and aggressive measures for those high risk patients.
Prediction and Tracking
Artificial Intelligence is also used in developing mathematical models to study the transmission rate of COVID-19. The different mathematical models that are used include: (1) SIR( Susceptible, Infectious and Recovered) model (ii) GLEaM( Global Epidemic and Mobility) model (iii) TRANSIMS(Transportation Analysis and Simulation) system and (iv) IBM ( Individual Based model).
These models can theoretically predict the number of positive cases and the rate of transmission for the COVID-19 pandemic. Artificial Intelligence can easily identify the most vulnerable regions by tracking the number of confirmed cases and take necessary actions to curb the spread.
SIR models the spread of the disease and infection rate among isolated population. This model can effectively explain why some countries are able to control the spread of the infection while others are still struggling. GLEaM and TRANSIMS both account physical contact patterns due to travel. TRANSIMS even uses real-time modelling by placing sensors in outbreak regions.
A group of MIT professors are developing models to restart the economy. The project involves analyzing the risk of infection, hospitalization and death of different age groups. This model will allow to restart the economy while simultaneously saving lives of the senior citizens that are more likely to be infected.
Spreading of Misinformation
AI is also used to response to this crisis by controlling misinformation. Social media is now actively using personalized AI technology to mitigate the spread of false information across their platform. This becomes particularly important because a series of conspiracy theory and false medical information are circulating across the media. AI can make the process faster for screening of false information regarding COVID-19.
Development for drugs and vaccines
It can help detect useful drugs that can be used to treat COVID-19 patients. AI can allow vaccines and treatments to be identified at a faster rate through computational analysis that can be helpful for clinical trials.
IBM Watson Health uses a database for determining the right medication for a patient. This is important because patients can have many underlying conditions and AI technology can actively identify the probabilistic medication ideal for the patient.
Artificial Intelligence can also be used to develop the vaccine and treatment drug. The process is usually labor intensive because of the selection of doses of various drug combination which includes a series of trials and errors. AI can reduce this cost and make the process more efficient.
IDENTIF.AI is an AI platform that is used to pinpoint Remdesivir i combination with Ritonavir and Lopinavir as a regimen against SAR-COV-2. However this theory is still not peer reviewed but offers optimism of the technology for future success in controlling COVID-19.
MIT is also collaborating with IBM researchers to develop an AI tool to assist physicians to find ventilator settings that will determine how long a patient needs to kept in the machine. This is particularly important because shortened ventilation time can reduce exposure to possible lung damage and allow scope for use by other intensive care patients.
There are several other relevant industrial and academic research on-going to help us curb the spread of this infectious disease.
AI has shown success in early diagnosis, tracking, controlling the spread of misinformation and also development of potential drugs and vaccines. It is not only during the pandemic but it has always complimented the work of the healthcare industry. It is one of the few resources that can allow cost effective, less time consuming method to develop vaccines and treatment to help us fight SARS-COV-2.
Vaishya, R., Javaid, M., Khan, I. H., & Haleem, A. (2020). Artificial Intelligence (AI) applications for COVID-19 pandemic. Diabetes & Metabolic Syndrome: Clinical Research & Reviews.
Chinazzi, M., Davis, J. T., Ajelli, M., Gioannini, C., Litvinova, M., Merler, S., … & Viboud, C. (2020). The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science, 368(6489), 395–400.
Volpert, V., Banerjee, M., & Petrovskii, S. (2020). On a quarantine model of coronavirus infection and data analysis. Mathematical Modelling of Natural Phenomena, 15, 24.
Balcan, D., Colizza, V., Gonçalves, B., Hu, H., Ramasco, J. J., & Vespignani, A. (2009). Multiscale mobility networks and the spatial spreading of infectious diseases. Proceedings of the National Academy of Sciences, 106(51), 21484–21489.
Eubank, S., Guclu, H., Anil Kumar, V. et al. Modelling disease outbreaks in realistic urban social networks. Nature 429, 180–184 (2004). https://doi.org/10.1038/nature02541
Degli Atti, M. L. C., Merler, S., Rizzo, C., Ajelli, M., Massari, M., Manfredi, P., … & Iannelli, M. (2008). Mitigation measures for pandemic influenza in Italy: an individual based model considering different scenarios. PloS one, 3(3).
IDentif.AI: Artificial Intelligence Pinpoints Remdesivir in Combination with Ritonavir and Lopinavir as an Optimal Regimen Against Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). Agata Blasiak, Jhin Jieh Lim, Shirley Gek Kheng Seah, Theodore Kee, Alexandria Remus, De Hoe Chye, Pui San Wong, Lissa Hooi, Anh T.L. Truong, Nguyen Le, Conrad E.Z. Chan, Rishi Desai, Xianting Ding, Brendon J. Hanson, Edward Kai-Hua Chow, Dean Ho. medRxiv 2020.05.04.20088104; doi: https://doi.org/10.1101/2020.05.04.20088104
This post was originally published by Shadeeb Hossain at Towards Data Science