Using advanced processing techniques on raw thermal images, researchers at the Georgia Institute of Technology (Georgia Tech) accurately tracked vital signs in a subject, including the subject’s heart rate, respiration rate, and body temperature. To enhance the clarity and quality of the thermal images, the team used phasor thermography, a technique it developed for hyperspectral, high-resolution, multiparametric thermal imaging and vision. Phasor thermography enables passive, contactless, reliable, and detailed measurement of vital signs. In the future, it could be used to identify subtle changes in the body that could be early indicators of cancer and other diseases. Conventional thermal imaging typically does not make sharp differentiations between slight temperature variations. Also, heat in the environment can make the images too noisy to precisely measure physiological signals. The new technique overcomes the spectral ambiguity inherent in conventional thermal imaging, sharpening the texture and detail that can be extracted from images and removing the effects of heat from the environment surrounding a subject. “With this phasor thermographic technology, we can enhance the accuracy and efficiency of thermal imaging to detect abnormalities,” researcher Dingding Han said. “Phasor thermography has the capability of getting material segmentation, which is not possible with only pure thermal imaging.” Researcher Dingding Han adjusts a thermal camera capturing an image of researcher Corey Zheng. Using an advanced processing technique on the raw thermal image, the researchers can accurately measure body temperature, heart rate, and respiration rate. The noncontact technology could open new possibilities for vital sign monitoring and early disease detection. Courtesy of Georgia Institute of Technology/Candler Hobbs. Using a series of filters, the researchers captured ten images of different parts of the long-wavelength IR spectrum, where thermal radiation is detected. Once they had the images, they used a mathematical tool borrowed from signal processing, called thermal phasor analysis, to analyze the patterns of heat in the images. The team’s use of full-harmonics phasor energy and high-order thermal phasor perception led to enhanced texture extraction and material classification. The team developed algorithms to resolve textures in 3D and in sizes smaller than one millimeter. This level of detail enabled the team to accurately distinguish slight variations in the thermal images of a human subject, such as the differences between the hair on the scalp and on the eyebrows, and the metal rims of the subject’s eyeglasses. Phasor thermography allowed precise unmixing and high resolution of the physical attributes essential to a thermal scene, diminishing the potential for inadequate temperature identification due to conventional decomposition directly from the total thermal radiation. The researchers demonstrated the phasor thermography system with phantom and living subjects in room-temperature settings. They verified the system’s robustness and reliability in detecting physiological signals such as body temperature, respiration rate, and heart rate across different body regions. The system was able to differentiate vital signs in scenes with multiple people and accurately capture variations in respiration rate before and after exercise. Phasor thermography is robust against complex and non-uniform environmental radiation sources, and is compatible with all major IR thermography platforms. It uses commonly available equipment, which makes it easy to adapt the technique to different scenarios. Han said the system could easily integrate into hospital and other healthcare settings. “We used a thermal camera and the filters to get the hyperspectral image data, so it’s scalable,” she said. “You could integrate this setup into virtually any thermal imaging platform.” The team plans to further develop the prototype system and work with physicians to apply phasor thermography to the detection of breast cancer. “Thermography could give us an advantage in early detection, because it could noninvasively detect abnormal cell activity that indicates early cancer,” Han said. “For example, tumor cells need more oxygen to reproduce, so their temperature will be a little bit higher than normal tissue. With this phasor thermography approach, we could spot that. “This could be a cornerstone for future broad biomedical diagnosis,” she said. “This can be the first step for the next generation of biomedical thermography for early detection and diagnosis of cancer...it’s the first prototype, with an ultimate goal of evolving the next versions and making it easier to use in hospitals and clinics.” The research was published in Cell Reports Physical Science (www.doi.org/10.1016/j.xcrp.2025.102501).