Human Factors

The emergence and spread of infectious diseases, like COVID-19, are on the rise. Can you identify patterns between population density and COVID-19 cases and identify factors that could help predict hotspots of disease spread?

BIOT

Summary

Brazil is one of the countries that least performs diagnostics for COVID-19. Identifying cases of fever is one way to monitor the incidence of COVID-19. Thus, the BIOT seeks to integrate and monitor people's body temperature on a large scale using thermal cameras and artificial intelligence to protect their privacy. BIOT data can indicate outbreaks of contamination and the flow of infected individuals, mitigating the problem caused by lack of diagnostics.

How We Addressed This Challenge

Fever is a common symptom of several infectious diseases, such as COVID-19 [1]. With the lack of sufficient diagnostics, thermal cameras can be used to detect outbreaks of Sars-CoV-2 infections in Brazil [2, 3]. With the help of artificial intelligence, BIOT can make this job more accurate, secure and highly integrated.

How We Developed This Project

The motivation:

Brazil already has almost 500 thousand individuals contaminated by the virus Sars-CoV-2 [4]. This picture gets worse with the gradual reduction of social isolation policies. Nevertheless, the authorities already recognize that the number of diagnostics performed is insufficient to monitor the progress of the pandemic in Brazil [5]. With this lack of diagnostics, controlling the outbreaks of COVID-19 and preserving people's lives becomes an even more difficult task [6].

The BIOT is a system thought by the United Cheese Bread team that integrates thermal cameras and artificial intelligence. Our thermal cameras identify only the body temperature and do not use facial recognition or GPS, preserving people's privacy. The data generated by the BIOT can be used to identify contamination hotspots and assist in decision making when diagnoses are not sufficient.


Why is people's temperatures so important?

  • In COVID-19, the fever can be found in 99% of sick individuals, with an average incubation period of 4 to 5 days [1].
  • In the case of MERS-CoV, the body temperature can occur in up to 98% of sick individuals, with an average incubation period of 2 to 14 days[7].
  • Fever is one of the first symptoms of severe acute respiratory syndrome [8].
  • In the case of H1N1 the body temperature was found in 94% of 642 patients who participated in a study, with a mean incubation period of 1 to 4 days [9].

Fever is also a good parameter to search in local outbreaks, such as those occurring in prisons, where there is a higher rate of infectious diseases such as Hepatitis B, Hepatitis C and H1N1, or in schools, where upper airway and gastrointestinal infections are more frequent.


Project architecture:

Initially it is necessary to choose the strategic places of the city (preferably with large flow of people) where you want to monitor. Then, based on the desired locations, one of the two types of monitors is used:

  • The first type is with a smartphone set and a thermal camera, where an application with artificial intelligence detects people's body temperature and sends the data to the system that saves in the database (see the link below).
  • The second type is with a set that combines a surveillance camera, a thermographic camera, an NVR module and an internet radio antenna. This set sends the images in real time to the system that uses artificial intelligence to analyze people's body temperature and save the data obtained in the database.

When starting its operation, BIOT can provide data in real time while respecting the privacy of individuals. The BIOT system interface can be visualized in the following link:

https://drive.google.com/file/d/1Rb_MXHFw4byMsRMqExmTXQZ6ZTBD5jUq/view


The data that will be obtained:

With the monitoring of the BIOT system, it is possible to obtain data and create estimates like:

  • The number of people with fever who passed through each camera.
  • The average body temperature of individuals in each region.
  • The estimate of areas with the highest rate of infections.
  • The number of individuals analyzed per day and in each region.

Through these data it is possible to verify the places with peaks in the number of individuals with fever. Thus, the authorities can monitorate and be alert with possible outbreaks of infection.


How to maintain people's privacy?

To maintain people's privacy, we choose not to do facial recognition. The artificial intelligence used in BIOT will recognize who is a person and will determine his/her body temperature, so that only the temperature is saved in the database. Therefore, all monitoring data saved in the database is anonymous and it is impossible to track specific people.


Use case:

The city of São Paulo has more than 12 million inhabitants, being the most populous city in Brazil [10]. The subway and train lines of the city of São Paulo transport about 7 million people a day [11]. Once installed in São Paulo's subways and trains, the BIOT system would be able to monitor the body temperature of a large portion of the population daily.

Data & Resources

1. Özdemir Ö. Coronavirus Disease 2019 (COVID-19): Diagnosis and Management. Erciyes Med J 2020; 42(3): 00–00. doi:10.14744/ etd.2020.99836.

2. Quilty BJ, Clifford S, Flasche S, Eggo RM; CMMID nCoV working group. Effectiveness of airport screening at detecting travellers infected with novel coronavirus (2019-nCoV). Euro Surveill. 2020;25(5):2000080. doi:10.2807/1560-7917.ES.2020.25.5.2000080

3. Naudé, Wim, Artificial Intelligence Against Covid-19: An Early Review. IZA Discussion Paper No. 13110. Available at SSRN: https://ssrn.com/abstract=3568314

4. COVID-19 Coronavirus Pandemic. Available in: https://www.worldometers.info/coronavirus/ Access in: 05/30/2020.

5. Ministério admite que quantidade de testes para coronavirus no Brasil não é ideal. O Estadão. Available in: https://saude.estadao.com.br/noticias/geral,ministerio-da-saude-admite-que-quantidade-de-testes-para-coronavirus-feita-no-brasil-nao-e-ideal,70003315597 Access in: 05/30/2020

6. Testes para coronavírus no Brasil não chegam a 10% da taxa dos EUA. Estado de Minas. Available in: https://www.em.com.br/app/noticia/gerais/2020/05/29/interna_gerais,1151761/testes-para-coronavirus-no-brasil-nao-chegam-a-10-da-taxa-dos-eua.shtml Access in: 05/30/2020.

7. Choi WS, Kang CI, Kim Y, et al. Clinical Presentation and Outcomes of Middle East Respiratory Syndrome in the Republic of Korea. Infect Chemother. 2016;48(2):118‐126. doi:10.3947/ic.2016.48.2.118

8. Kaydos-Daniels SC, Olowokure B, Chang H, Barwick RS, Deng J, Lee M, et al. Body Temperature Monitoring and SARS Fever Hotline, Taiwan. Emerg Infect Dis. 2004;10(2):373-376. https://dx.doi.org/10.3201/eid1002.030748

9. Novel Swine-Origin Influenza A (H1N1) Virus Investigation Team, Dawood FS, Jain S, et al. Emergence of a novel swine-origin influenza A (H1N1) virus in humans [published correction appears in N Engl J Med. 2009 Jul 2;361(1):102]. N Engl J Med. 2009;360(25):2605‐2615. doi:10.1056/NEJMoa0903810

10. Aos 466 anos, São Paulo tem 11,8 milhões de habitantes. Agência Brasil. Available in: https://noticias.r7.com/sao-paulo/aos-466-anos-sao-paulo-tem-118-milhoes-de-habitantes-20012020 Access in: 05/30/2020.

11. CPTM e Metrô: novas estações e 7,8 milhões de usuários por dia. Portal do Governo. Available in: https://www.saopaulo.sp.gov.br/ultimas-noticias/cptm-e-metro-novas-estacoes-e-transporte-de-78-milhoes-de-pessoas-por-dia/ . Access in: 05/30/2020.

Tags
#IoT #artificialintelligence #covid
Global Judging
This project was submitted for consideration during the Space Apps Global Judging process.