A cheaper, faster and safer way to filter temperatures

Newswise – New York, NY – February 22, 2022 – It’s human nature to dislike waiting in line and the COVID pandemic has only made lines worse, especially in tall buildings where it can take hours to have your temperature checked before being allowed in Researchers from Columbia Engineering and the Mailman School of Public Health to have invented a system capable of automatically raising the temperature of passers-by, going about their business, up to three meters away, no one needs to stay a few seconds in front of a camera to take a measurement. And no one needs to be there to read the measurement and approve the person’s entry.

“Using advanced algorithms and models, we have developed an inexpensive way for hospitals, restaurants, subway stations, schools, high-rise buildings with elevators, etc. to perform fever screening without disrupting the normal flow of traffic, helping to restore some amount of ‘normal’ to our daily lives,” said the team leader Fred Jiangassociate professor of electrical engineering, who came up with the idea after seeing how his wife’s hospital was overwhelmed during the pandemic with the number of people coming through the entrance. “A nurse had to scan everyone nearby, and that didn’t ensure that everyone who passed by didn’t have a fever, because people can take fever suppressants.”

Jiang is an expert in intelligent embedded systems and their applications in mobile and wearable computing, smart built environments, Internet of Things and connected health applications. He teamed up with Andrew RundleProfessor of epidemiology, to explore ways to improve the accuracy of inexpensive thermal imagers. There are currently two methods for performing non-contact fever screening: non-contact infrared thermometers like a “thermometer gun” and infrared thermography systems, or “thermal cameras.” These approaches require people to stand close to the device for a few seconds, interrupting the flow of traffic, or are very expensive, preventing wide adoption. Jiang’s expertise in using low-cost sensors combined with signal processing and machine learning algorithms, and Rundle’s vast knowledge of epidemiology led the team to create SIFTER, a low-cost system based on an RGB thermal camera for continuous fever screening in multiple people.

SIFTER has three components: a sensor node, a cloud server, and a web-based user interface. It uses a smaller number of low-cost sensors, can operate continuously in different environmental conditions, and can screen multiple people simultaneously without requiring the active participation of screeners. The system detects and tracks heads in the RGB and thermal domains, builds thermal heat map models for each tracked person, and classifies people as having or not having a fever. It can take the main temperature characteristics of the heads on the spot from a distance of up to three meters and produce real-time fever screening predictions, dramatically improving screening throughput while minimizing disruption to normal activities.

The team tested SIFTER in the real world, for 10 months in conjunction with ColumbiaDoctors at their downtown New York clinic, with support from Teresa Spada, Director of Practice Operations at ColumbiaDoctors Midtown. He examined more than 4,000 people, with a measurement error of less than 0.4◦F at two meters and about 0.6◦F at 3.5 meters. By comparison, most infrared thermal scanners on the market, which cost several thousand dollars, have a measurement error of about 1◦F when measured at less than 0.5 meters. SIFTER can achieve a 100% true positive rate with a 22.5% false positive rate without requiring human interaction, significantly outperforming the baseline[6]which has a false positive rate of 78.5%.

SIFTER’s embedded software is open-source and the hardware costs less than $500 to install and operate. The data processing component is provided free of charge by Jiang’s Columbia Smart and Connected Systems Laboratory as a cloud service in containerized environments, allowing anyone to deploy this inexpensive solution anywhere in the world. http://icsl.ee.columbia.edu/FeverScreening/

The entire end-to-end system is fully automated, requiring no one to manually take measurements or analyze results. The researchers built SIFTER using a FLIR One Pro thermal camera and an embedded Jetson Nano processor to keep their costs down. The system uses signal processing, computer vision and deep learning to identify “features” or relevant parts of a person’s face, detect individuals’ heads, track their movements and estimate the orientation of the face. head using RGB images and thermal cameras. Then, based on head orientation and thermal camera images, the system extracts the temperature of uncovered areas of skin and maps them onto a 3D point cloud model of each person’s head. The system takes these temperature readings from different parts of the face and uses machine learning models to estimate the person’s body temperature and determine if they have a fever.

To manage the distance between people and the scanner, the researchers first estimated the distance between a person’s head and the camera, based on a perspective projection (closer objects appear larger than more distant objects) and stereo vision (perception of depth based on the distance between RGB and thermal cameras). Next, they modeled the physical transmission of heat from a person’s skin to the camera as a function of a person’s distance and used this to calibrate and remove noise from phenomena such as attenuation (the longer a force travels through a medium, the more it is attenuated and the weaker the reading you will receive). Finally, to further improve the efficiency of the system, the team combined physical modeling with data-driven machine learning methods to further improve and account for falling remote temperature readings.

Jiang and Rundle are currently working to improve SIFTER in terms of range and accuracy, so the system can be deployed in more scenarios, such as large open spaces and transportation hubs. They are also expanding their system to create a “Global Influenza and Emerging Disease Visual Surveillance System,” which they will jointly manage. The team plans to use this system as a public health surveillance tool to detect emerging infectious diseases and influenza, which continuously monitors the number of people in public while running a fever. When the system detects that the prevalence of people with fever has exceeded a normal low background rate, public health authorities will be alerted to initiate more thorough infectious disease surveillance and detection protocols.

“We are very excited about SIFTER,” Rundle added. “Fever is a symptom of many infectious diseases, and this system will provide a low-cost, near real-time, and continuous means of detecting spikes in the number of people with fever. We see SIFTER as a complement to other more expensive and slower public health surveillance systems that rely on laboratory testing and case reporting. We expect this system to help improve the detection of emerging infectious diseases and seasonal influenza outbreaks. ”

The study was accepted at 21st ACM/IEEE Conference on Information Processing in Sensor Networks (IPSN 2022), from May 4 to 6, 2022, in Milan, Italy.


About the study

The study is titled “A low-cost, in-situ system for continuous fever screening in multiple individuals.”

The authors are: Kaiyuan Hou, Yanchen Liu, Peter Wei, Chenye Yang, Hengjiu Kang, Stephen Xia and Xiaofan Jiang (all of Columbia Engineering); Andrew Rundle (Mailman School of Public Health); and Teresa Spada (Columbia University Irving Medical Center).

This research was funded by Columbia Engineering under the Technology Innovations for Urban Living in the Face of COVID-19 grant. Students working on this project were partially supported by the National Science Foundation under grant numbers CNS-1704899, CNS-1815274, CNS-11943396, and CNS-1837022.



SIEVE : http://icsl.ee.columbia.edu/FeverScreening/

Paper: http://icsl.ee.columbia.edu/publications/





Comments are closed.