The following translation, excerpted from a Chinese government think tank report on China’s use of digital health technology to fight COVID-19, describes several examples of how PRC AI companies are using AI technology to improve COVID-19 prevention and treatment. Most of the examples, such as meal delivery robots for hospital staff and technologies that allow doctors to diagnose patients at a distance, are benign. But a number of these AI applications raise troubling questions about privacy and human rights, such as tracking of mobile phones used by suspected infected individuals and video surveillance programs that can identify individuals by posture.
The Chinese source text is available online at: http://www.caict.ac.cn/kxyj/qwfb/ztbg/202003/P020200324369446692496.pdf
Translator: Etcetera Language Group, Inc.
Editor: Ben Murphy, CSET Translation Lead
III. Artificial Intelligence (AI) Example Applications Case 1: The Xiamen COVID-19 Tracing and Monitoring System: Opening a New Front in the Battle Against the Epidemic
Summary: Xiamen’s COVID-19 (新冠肺炎) tracing and monitoring system was designed and built in strict adherence to the principle of “externally, defend against the entry of new cases; internally, defend against spreading,” focusing closely on the actual needs of collection, case-by-case investigation, reporting, treatment, analysis, and application. Using big data integration and multi-dimensional analysis, an epidemic prevention and control model was constructed, key groups of concern were pinpointed, and sources of infection were blocked. It helped the epidemic prevention and control departments properly manage key areas, groups and situations, and provided strong decision support data for directing epidemic prevention and control at all levels, thereby achieving comprehensive control over epidemic prevention work citywide.
Keywords: big data, AI, digital prevention and control, COVID-19, knowledge graph
Excerpted Translation
Main Text:
Beginning in December 2019, with the growing spread of the novel coronavirus (新型 冠状病毒) infection, the COVID-19 control and prevention situation was becoming increasingly severe. In accordance with the unified arrangements of the Xiamen Novel Coronavirus Infection Control Headquarters, under commission by the Xiamen Municipal Sanitation and Health Commission and focusing on the work objective of “externally, defend against the entry of new cases; internally, defend against spreading,” the Xiamen Health and Medical Big Data Center researched and developed the Xiamen COVID-19 Tracing and Monitoring System.
The Xiamen COVID-19 Tracing and Monitoring System focuses closely on the actual needs of collection, case-by-case investigation, reporting, treatment, analysis, and application. Using big data integration and multi-dimensional analysis, an epidemic prevention and control model was constructed, key groups of concern were pinpointed, and sources of infection were blocked. It helped the epidemic prevention and control departments properly manage key areas, groups, and situations, and provided strong decision support data for directing epidemic prevention and control at all levels, thereby achieving comprehensive control over epidemic prevention work city-wide.
Problems addressed in the example:
1. Difficulty of case-by-case investigation and prevention and control for key groups
Grassroots case investigators have limited technical means and power at their disposal, and epidemic-related data and information fed back from various parties is complex. When following-up on close contacts with confirmed cases, for example, there is information from multiple organizations such as the Public Security Bureau, Public Transportation Group, Didi, and others, and analysis and comparison presents considerable 1 difficulty. It is not possible to establish real-time, automatic linking channels for data, and it is labor-intensive. Given the difficulty of data analysis and comparison, it is difficult to track and locate key individuals.
2. Difficulty tracking the process of dealing with hospital fever patient groups
With respect to information on people visiting the fever outpatient departments of the city’s hospitals, the relevant regulatory authorities were unable to record the handling and whereabouts of suspected cases of fever, making it impossible for community grid (社 区网络) personnel to monitor fever patients that doctors require to be under home observation. This results in problems in terms of being unable to do real-time tracking of case-by-case investigation, testing, observation, and treatment of individuals with fevers.
3. Difficulty tracing close contacts
In traditional infectious disease investigation, information on close contacts is obtained through manual questioning, relying on patients’ recollection of possible contacts, places visited, transportation used, etc. The large amount of information, vague memories, and other factors, however, result in the omission of important information, and erroneous information provided by patients may cause the workloads of epidemiological investigators to increase, and compound the difficulty of their work. How to ensure that epidemiological investigators can smoothly carry out infectious disease investigations, quickly locate close contacts, promptly arrange isolation and checking, and stop the spread of epidemics, has become an important question.