Skin imaging is a powerful, noninvasive method used with potential to aid in the computer-assisted diagnosis of numerous dermatological diseases and assess overall skin health. By tracking the evolution of various skin features, we can monitor skin health. One interesting feature is known as the “microrelief,” which are the fine micrometer scale furrows and ridges that appear like irregular geometric patterns on the skin surface. However, it is difficult to accurately observe the microrelief structure and its evolution over time due to the micrometer dimensions of the microrelief and the 3D non-rigid nature of the body. Registration and matching of the same skin region are further complicated by noisy and distorted optical images. We have designed a high resolution, handheld optical system to image the skin microrelief. The device has potential to be used in clinical settings since it is small and lightweight. With proper experimental design, we are able to acquire repeatable images of a selected skin patch to monitor over time. Additionally, we have developed methods for registration of skin patches and analyzing skin feature stability. Using real and synthetic skin images, we demonstrate that we can accurately and robustly register large area skin images and identify skin pattern correspondences. Essentially, through repeatable, high resolution imaging, we can monitor the microrelief structure in select individuals over a period of 1-2 years. This has interesting applications because we can use the microreliefs for health monitoring and as a map for the body since we notice that these features are stable over time in healthy individuals.
The human body is comprised of a variety of networks that can be monitored and used as body positioning systems. Furthermore, structural changes observed in these networks have clinical significance as they can aid in disease diagnosis and determining the overall health of an individual. One such network is the superficial vascular structure. As the primary network supplying blood to the body, observing the vein structure gives insight into the cardiovascular health and hydration levels of an individual. Additionally, because of the uniqueness of the network, there is growing interest in using veins as a biometric for identification and mapping. However, because vasculature is difficult to image and existing imaging technology is expensive, the potential for superficial vascular structure to map the body and provide insight into overall health has not been well studied. Furthermore, given the 3D nature of the body, registering and matching corresponding vascular regions proves to be quite challenging. In order to address these needs, we have designed a near-infrared (NIR) imaging system to image the superficial vascular structure. It is compact, easily integrated into any computer system, and cost-effective, thereby having the potential to be used in clinical settings. By carefully designing the image acquisition system and developing registration and matching algorithms, we can robustly image and extract the vascular structure. By extracting the vascular structure from certain limbs, we show the potential for using vasculature as a body map. Additionally, we demonstrate the uniqueness of the vascular structure and its potential to be used as a biometric identifier.