ThermaiScan - COVID-19 live screening

Idea Visualisation

The world was not prepared for the current COVID-19 pandemic, which rapidly spread to all parts of the world. Evidently we did not have sufficient health security measures in place, compared to physical security infrastructure that was enhanced after 9/11. We and many others believe that governments post-COVID-19 will impose further health security measures to cope with current or future virus spreads. COVID-19 has revealed a greater need for a health security perspective to protect risk groups whilst keeping the health care system, factories & businesses up and running. We see this need for workers and visitors in warehouses, gym’s various distribution centres and commercial buildings. We believe that needs in health care will change post COVID-19 towards more health security in hospitals and similar facilities. Many mini clinics will pop up or expand for pre-screening and telemedicine will most likely grow, meaning new methods of screenings are needed.

Our screening system combine two different cameras and three data sources: Infrared thermography (IRT) camera absorbs heat of facial skin temperature & respiratory signal data. IRT detects the temperature change around the nasal area, estimates the respiratory rate and measures the facial skin temperature. Charged-Coupled Device (CCD) cameras that capture the heartbeat data. The CDD camera detects even the smallest color changes of facial skin induced by blood flow. Analysis of the screening is powered by different types of Machine Learning algorithms that we will develop that informs if someone is healthy or has a risk of infection. Our method is based on combined three vital-sign-based screenings; facial skin temperature, respiratory signal and heartbeat signal rates. Infections have a direct association with inflammation, which causes elevations in facial skin temperature, heart and respiration rates. By using these three vital-signs, the detection accuracy of the system improves from 81.5 to 98.0% in case-control studies. This is notably higher compared to the conventional screening methods using only infrared thermography (IRT). Infection screening detection accuracy can further be increased by applying Machine Learning and image processing algorithms to the analyses of these vital signs. The screening is non-invasive and does not have any type of radiation that can harm. We estimate the project will be completed within 6-12 months.

The potential of guaranteeing virus free facilities can naturally have big positive outcomes and solve a real need in various areas in society. Businesses, governmental and private facilities can be kept open and operating, thus minimizing the negative impact on health and economy COVID-19 is causing to the world nowadays. We have a great opportunity to gather data from infection clinics and hospitals for our Machine Learning development during this pandemic period and therefore the screening can become better and better. Our solution and technology will be developed with our pilot customer Erebuni Medical Center. First we develop the MVP (Minimal Viable Product) and then collect data for improving and validating our algorithms. The next stage is to test our innovation via clinical trials together with more pilot customers in different industries. When we have a functioning product and proof of concept we will share our innovation globally.

Developing actor: ThermaiScan is a Swedish Healthtech startup focusing on predictive medicine based on AI & Thermal Technology. Thermaiscan is the organizer and owner of this project. We have expertise in Infrared thermography (IRT) & Charged-Coupled Device (CCD) technology and Machine Learning in our team. Need owner: Erebouni Medical Center in Yerevan, Armenia has the need to make their hospital infection / virus free and has the capacity to co-create and implement our solution. We also have access to Nairi Medical Center in Yerevan and Dilijan Medical Centre in Dilijan city, Armenia. It is the same owner for all three hospitals.

The risk we see during developing the prototype and for the project overall is access to COVID-19 patients. It's high risk and the consequence will be that it slows down progress invalidation of algorithms. So that is why we need to be proactive and have more than one data collection center (Infection clinic/hospital) ready in our pipeline. Another risk is individual integrity and privacy. It’s crucial that our innovation is used where the individual's data integrity and privacy are respected and at the same time protecting the company / the collective by creating safe environments. We understand that it will be case by case how we implement our solution in hospitals, elderly homes, and factories, some places are more manual and some places automated with gates/doors. Together with our customers and the people that will be scanned we will create the screening journeys so that it satisfies all involved parts. Issues and friction discovered during the pilot customer period will be iterated. We recognize that our screening solution is only part of solving the problem with the infections/viruses. By finding viruses spread early with risk assessments in strategic places, as an example, the health care system can combine our screening method with other services to prevent further spread.