Thermal signatures of skin lesions using computational thermal modeling and medical infrared imaging
Johns Hopkins University
Infrared (IR) thermography is a valuable quantitative diagnostic tool that allows for non-invasive, accurate measurement of skin temperature variations in the presence of a lesion. Modeling the underlying thermal and physiological processes within the body offers excellent potential for improving the thermographic measurement system design and developing more exact, quantitative assessment criteria. Using computational modeling and infrared imaging experiments, this dissertation investigates the thermal signatures of lesions of varying geometrical and physiological characteristics. We first performed a comprehensive sensitivity analysis of the computed skin temperatures in order to understand the relationships between healthy skin temperatures and the underlying thermophysical processes and tissue properties. These functional relationships provide a foundation for interpreting steady state and transient thermal signatures of skin lesions. We developed a computational thermal model for a heel deep tissue injury (DTI) to allow for an early thermographic detection and assessment capability for DTIs. The DTI models were used to develop thermographic measurement strategies and quantitative staging criteria that can be employed in a clinical setting. We analyzed the infrared images of various vascular tumors and pigmented skin lesions acquired from patient studies, using the combined white light-infrared image processing approaches. Our quantitative thermal analysis of lesions of different physiological characteristics, sizes, locations and depths will facilitate quantitative assessment and interpretation of other skin lesion thermographic images. A better understanding of the thermal behavior of skin lesions, gained using computational modeling and infrared imaging experiments in this study, can contribute to the advanced use of quantitative infrared imaging in medical diagnostic applications.
Skin lesions, Infrared (IR) imaging, computational model, bioheat transfer