SARS-COV-2-19 Diagnosis Through Integration of Thermal Video and Patient Data

Abstract

INTRODUCTION: The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic rapidly spread across the world, creating a need for real-time screening of individuals to prevent further transmission. We aimed to develop a mobile thermographic imaging platform, intended to scan an entrance or hallway to identify people with abnormal facial temperatures and/or vital signs.

METHODS: Adult patients who were actively symptomatic with SARS-CoV-2 infection were included along with people with other febrile illnesses, such as pneumonia or viral infection, and healthy/asymptomatic individuals, for comparison. We applied an attention-based model to conduct a thermal image-based diagnosis of the disease. Thermographic images and videos of the study subjects were obtained through a hand-held thermal camera at distances of 5 and 10 feet. Subsequently, the videos were processed and analyzed in the Photogrammetric Computer Vision Laboratory to attempt to categorize individuals with abnormal facial temperatures and vital signs.

RESULTS: Out of a total of 76 participants, 35 patients were positive for SARS-CoV-2 with active infectious syndromes, 21 were diagnosed with other infectious diseases, and the remaining 20 were asymptomatic control subjects. The dataset consists of short videos recorded by the thermal camera to indicate the temperatures of the participants. 53 videos were obtained for SARS-CoV-2 patients and 22 for non-SARS-CoV-2 patients, where 15 frames are sampled from each video at an equal distance, and each frame is resized as 224 pixels.

CONCLUSIONS: This study generated preliminary methods and results aimed at the early identification of febrile illness and can be used as a baseline for future research.