Thermo-regulatory response of static and dynamic breast thermography as a biomarker of malignancy

Abstract This study assesses the diagnostic potential of static and dynamic infrared thermography by relating physiological vascular patterns to established indicators of breast malignancy. Mammography remains the principal global screening modality; however, its accuracy declines in women with dense breast tissue [1], and its high cost limits access in developing countries [2]. Although dynamic thermal imaging may improve lesion detection, its clinical implementation in developing settings is constrained by the opacity of artificial intelligence models and the limited availability of high-quality datasets [3]. This paper presents a physiology-based model intended to reduce reliance on extensive training data and thereby facilitate analysis in small patient cohorts. Earlier Virtual Wave Transformation approaches appear to capture dynamic thermal responses but struggle to differentiate metabolic heat from vascular signals. The present study combines dynamic thermal imaging with static infrared imaging. By exploiting the different temporal behaviors of these signals, the model may distinguish increased cancer-related metabolic activity from other thermal patterns. In contrast to black-box deep-learning systems and even explainable AI approaches, this algorithm provides a more interpretable mathematical alternative by decomposing thermal patterns and separating convective vascular heat transport from thermal conductive metabolic signals. Analysis of the temporal variation in each component's amplitude indicates a possible association with patient diagnosis. Whereas conventional AI methods generally require large databases and expensive clinical validation, such as MRI, this approach uses physiological parameters to obtain diagnostic information from smaller samples. This work is a limited pilot study focused on evaluating the mathematical methods; we did not perform detailed statistical analysis.