FDA Approves Machine Learning Tool for COVID-19 Screening

The FDA has issued an emergency use authorization (EUA) for the first machine learning-based non-diagnostic screening tool to prevent the spread of COVID-19.  

The Tiger Tech COVID Plus Monitor is intended for use by trained personnel to help reduce exposure to and transmission of SARS-CoV-2, the virus that causes COVID-19. The tool identifies certain biomarkers that may be indicative of SARS-CoV-2 infection as well as other hypercoagulable conditions (like sepsis or cancer) or hyper-inflammatory states (like severe allergic reactions) in asymptomatic individuals over the age of five.

The Tiger Tech COVID Plus Monitor is designed for use following a temperature reading that does not meet criteria for fever in settings where temperature check is being conducted in accordance with the CDC and local institutional infection prevention and control guidelines.

The FDA noted that the tool is not a substitute for a COVID-19 diagnostic test and is not intended for use in individuals with symptoms of COVID-19.

“The FDA is committed to continuing to support innovative methods to fight the COVID-19 pandemic through new screening tools,” said Jeff Shuren, MD, JD, director of FDA’s Center for Devices and Radiological Health.

“Combining use of this new screening device, that can indicate the presence of certain biomarkers, with temperature checks could help identify individuals who may be infected with the virus, thus helping to reduce the spread of COVID-19 in a wide variety of public settings, including healthcare facilities, schools, workplaces, theme parks, stadiums and airports.”  

The device is an armband with embedded light sensors and a small computer processor. During use, the armband is wrapped around a person’s bare left arm above the elbow. The sensors first obtain pulsatile signals from blood flow over a period of three to five minutes.

Once the measurement is completed, the processor extracts some key features of the pulsatile signals, such as pulse rate, and feeds them into a probabilistic machine learning model that has been trained to make predictions on whether the individual is showing certain signals, like hypercoagulation in blood.

Hypercoagulation is known to be a common abnormality among COVID-19 patients, FDA stated. The result is provided in the form of different colored lights used to indicate if an individual is demonstrating certain biomarkers or if the result is inconclusive.

Researchers studied the clinical performance of the Tiger Tech COVID Plus Monitor in both hospital and school settings.

The hospital study, which was considered a validation study, enrolled 467 asymptomatic individuals, including 69 confirmed positive cases, and demonstrated that the Tiger Tech COVID Plus Monitor had a positive percent agreement (proportion of the COVID-19 positive individuals identified correctly by the device to possess certain biomarkers) of 98.6 percent.

Additionally, researchers found that the tool had a negative percent agreement (proportion of the COVID-19 negative individuals identified correctly by the device to not possess certain biomarkers) of 94.5 percent. In the school study, which was considered a confirmatory study, the tool showed a similar performance.

The FDA emphasized that the Tiger Tech COVID Plus Monitor is not a diagnostic device and must not be used to diagnose or exclude COVID-19 infection. The device is also intended for use on individuals without a fever. An individual’s underlying condition may interfere with the COVID-19 related performance of the device and could lead to an incorrect screening result.

The FDA has increasingly sought to enhance the efficient use of machine learning and AI in healthcare. In January 2021, the agency released its first AI and machine learning action plan, a multi-step approach designed to improve the FDA’s management of advanced software.

“This action plan outlines the FDA’s next steps towards furthering oversight for AI/ML-based SaMD,” said Bakul Patel, director of the Digital Health Center of Excellence in the Center for Devices and Radiological Health (CDRH).

“The plan outlines a holistic approach based on total product lifecycle oversight to further the enormous potential that these technologies have to improve patient care while delivering safe and effective software functionality that improves the quality of care that patients receive. To stay current and address patient safety and improve access to these promising technologies, we anticipate that this action plan will continue to evolve over time.”