Unpredictable and rapidly spreading, the Covid-19 coronavirus has travelled from its earlier concentration in China to reach more than 100 countries, infecting over 110,000 people and causing upwards of 4,000 deaths thus far. Given its worldwide impact, the World Health Organization last month declared the coronavirus a global health emergency.
In comparison to the last major outbreak of the SARS virus in 2003, those fighting this particular epidemic can leverage new and emerging technologies that quickly aid public health bodies in creating a valuable understanding of the coronavirus, guiding prevention efforts, augmenting human aid and support efforts, and facilitating virus research.
In recent years, AI has begun to play a significant role in the health-care sector: Advanced computing and data-analysis tools enable information sharing and diagnostic practices, and deepen the medical profession’s understanding of diseases and infections. Prompted by the urgent need to contain Covid-19, government agencies and private companies around the world are increasingly looking toward AI-based techniques to provide insight on its spread and support treatment for those who have been infected.
Because of Covid-19’s unpredictable but highly-contagious nature, academic and medical communities have prioritized analyzing the structure of the virus in order to create an effective vaccine. However, this virus is particularly challenging because it belongs to a family of enveloped coronaviruses that contain single-strand RNA structures. Similar to other double-stranded viruses including HIV, Ebola, Influenza, and others, Covid-19 is capable of rapidly mutating, making vaccine development and virus analysis difficult.
To support the research, Baidu has made its Linearfold algorithm available to scientific and medical teams fighting the outbreak. The Linearfold algorithm, published in partnership with Oregon State University and the University of Rochester in 2019, is significantly faster than traditional RNA folding algorithms at predicting a virus’s secondary RNA structure. Analyzing the secondary structural changes between homologous RNA virus sequences (such as bats and humans) can provide scientists with further insight into how viruses spread across species. Due to the recent outbreak, Baidu AI scientists have used this algorithm to predict the secondary structure prediction for the Covid-19 RNA sequence, reducing overall analysis time from 55 minutes to 27 seconds, meaning it is 120 times faster.
Recent studies show that the secondary structure of messenger RNA (mRNA), an RNA version that instructs cells to make proteins, is correlated to its functional half-life and has an impact on protein translation. Access to quick viral structural analysis can significantly shorten the time it takes to design a potential mRNA vaccine with higher stability and better effectiveness, providing an opportunity to save thousands of lives. Along with opening the Linearfold algorithm to the broader community, Baidu collaborates with health and academic institutions to share computation resources, provide customized support, and help optimize mRNA vaccine design.
“We hope that this powerful ability can be quickly leveraged by our researchers and anti-epidemic experts, and work with society as a whole to help improve the speed of virus research and vaccine development,” says Haifeng Wang, CTO at Baidu.
As the virus continues to spread, it is becoming increasingly important to provide quick monitoring tools, particularly in high-traffic public spaces such as transportation hubs, to effectively quarantine and reduce the spread of the coronavirus.
Baidu has developed several tools that are effective in building awareness and screening populations, including an AI-powered, non-contact infrared sensor system that provides users with fast multi-person temperature monitoring that can quickly detect a person if they are suspected of having a fever, one of the many symptoms of the coronavirus. This technology is currently being used in Beijing’s Qinghe Railway Station to identify passengers who are potentially infected where it can examine up to 200 people in one minute without disrupting passenger flow.
“Traditional approaches, such as station personnel using frontal medical thermometer devices, can easily cause crowds and increase the risk of cross-infection. Baidu's AI temperature sensor system can quickly screen crowds to improve detection efficiency and accuracy without creating unnecessary risks,” says Wang.
In addition, Baidu also released the industry’s first open-source model to detect whether individuals in crowded areas are wearing masks. The model boasts a classification accuracy of 97.27% with a robust performance in long-tail scenarios, such as oronasal mask or side faces. As the outbreak continues to spread, this tool also provides a key indicator of the population’s view and response to the coronavirus, by gauging the public’s adoption of safe health procedures like wearing face masks and regularly washing hands.
A CT scan of the chest is recognized as the primary diagnostic tool for pneumonia, a common effect of Covid-19. With limited front-line doctor resources to read an exponentially growing number of scans, fast and accurate CT imaging technology is critical to help clinicians detect and monitor the infections more effectively.
LinkingMed, a Beijing-based oncology data platform and medical data analysis company, released China’s first open-source AI model for pneumonia CT image analysis powered by Baidu’s open-source deep learning platform, PaddlePaddle. The AI model can quickly detect and identify pneumonic lesions while providing quantitative assessment for diagnosis information, including the number, volume, and proportion of pneumonic lesions.
By leveraging PaddlePaddle and the semantic segmentation toolkit PaddleSeg, LinkingMed has developed an AI-powered pneumonia screening and lesion detection system, putting it into use in the hospital affiliated with XiangNan University in Hunan Province. The system can pinpoint the disease in less than one minute, with a detection accuracy of 92% and a recall rate of 97% on test data sets.
In addition, many vulnerable populations are in areas with limited medical resources, making online access to care essential. Baidu’s online doctor consultation platform provides this essential service to those who want to consult medical professionals about Covid-19 without putting themselves or others at risk of getting infected. Currently, the platform has a massive network of doctors—over 100,000 specializing in respiratory systems and emergency room practices—and has already handled over 15 million enquiries from users, helping to raise awareness of the coronavirus’s many symptoms while providing patients with a custom care plan based on their severity.
Automated telecommunication services can also be used to share valuable information about the coronavirus to the public while tracking potential spread from a local perspective. As an example of how this has been applied to the coronavirus threat, Baidu has developed an intelligent robocall platform that has made over three million automated phone calls—1,500 calls in one second—requesting people voluntarily provide their recent travel history, close contacts, and current health conditions. Health commissions, provincial centers for disease control, and neighboring committees are able to use this platform to track the spread of the coronavirus across specific regions and promote medical preparedness from a local perspective.
AI and big data can be used to better understand key events and locations related to the coronavirus, pinpoint its origin, and measure the rate of spread. Such techniques can include using AI-powered data analytics to pull insights from online behavior such as online search queries and social media conversations to identify signals from a specific population that can provide insight as to the movement of the coronavirus.
The Boston Children’s Hospital used an automated HealthMap system that scans online news and social media reports for early warning signs of outbreaks, which led to the initial awareness that Covid-19 was spreading outside China. This initial alert spurred more detailed reports from other agencies including a warning from the Program for Monitoring Emerging Diseases (ProMed), a well-established volunteer-led organization, just half an hour after the first warning.
In response to the outbreak, Baidu has leveraged AI-powered mapping systems to identify the flow of travel across high-risk areas using Baidu Maps’ “Migration Big Data Platform.” The population movements out of Wuhan can broadly track the early spread of the coronavirus. AI is helping epidemiologists build an approximate picture of people’s migration with some carrying coronavirus. AI-driven analytics can provide users and health-care professionals with valuable real-time insight into the coronavirus’s spread that can accelerate local preparedness and response efforts.
Access to health care and resources at a moment’s notice is vital for battling the spread of the coronavirus. Autonomous vehicles are playing a useful role in providing access to necessary commodities for health-care professionals and the public alike by delivering goods in infected areas and disinfecting hospitals, effectively minimizing person-to-person transmission and alleviating the shortage of medical staff.
Apollo, Baidu’s autonomous vehicle platform, partnered with a local self-driving startup called Neolix to deliver supplies and food to the Beijing Haidian Hospital. With daily deliveries, this initiative is helping feed over 100 frontline staff members as they work to treat their growing patient base. Baidu Apollo has also made its low-speed driverless micro-car kits and autonomous driving cloud services available for free to companies dedicated to fighting the coronavirus.
With new coronavirus hotspots emerging in Italy and Korea, government agencies and companies like Baidu are using AI in a variety of settings to reduce peoples’ exposure and fast-track the virus’s containment. AI technology offers a versatile range of functions, allowing us to deeply understand and predict the evolution of the coronavirus, providing rapid insights to influence public perception, expediting the development of potential vaccines, and offering aid to those infected as well as the thousands of medical professionals committed to fighting it. This outbreak has also showcased the scale and power of digital connectivity: People across the globe are building networks, sharing information, and collaborating online at a scale that is unprecedented in the history of viral epidemics.