Calibrating Imaging Techniques to Study Blood Flow

This study is designed to calibrate three non-invasive and non-ionizing imaging techniques on 12 healthy volunteers. The three imaging techniques-- thermography, laser Doppler imaging and multi-spectral imaging-- have been approved since 2001 for four clinical protocols already approved by the NIH/NCI IRB for use on patients with Kaposi s sarcoma (KS). However, as our laboratory continues to study and analyze the images collected on these protocols, we have found that our analysis algorithms require some additional data from healthy volunteers.

We aim to use the information we gather from healthy volunteers on this protocol to train our imaging systems and calibrate our analysis methods to validate the results of the KS data already collected. We aim to study different skin types, such as Caucasian, Asian and African American, so that we can calibrate the melanin input value in our multi-spectral imaging reconstruction algorithm. We also need to study the vasculature networks in the forearms of the volunteers to compare to 'normal' values in the literature for the parameters we are exploring, including temperature, vasculature, blood volume and blood oxygenation to validate our reconstruction algorithm. We will also perform experiments of reactive hyperemia, where the arm of the volunteer is occluded with an arm pressure cuff for five minutes, to study how blood volume and blood oxygenation change during the experiment. Trends of increasing/decreasing blood volume and blood oxygenation can also be compared with published literature to validate our reconstruction algorithm.

Since we started collecting data from the KS patients, we have noticed that hair and curvature of the surface of the skin interfere with our analysis techniques as well. Therefore, we aim to assess algorithms developed to remove the effects of curvature and hair on the images as part of our image analysis training. The intensity in the images is affected by the curvature and must be corrected. We have developed algorithms to correct for this curvature, but need to study normal disease-free skin to make sure that the values for blood volume and blood oxygenation remain the same after the curvature correction is performed. We also plan to collect images from a healthy volunteer s arm, remove the hair from the arm using a topical hair removal solution, and then image the arm again. With this information, we can optimize our hair removal algorithm. Combining all of the aforementioned information will allow us to develop a non-invasive strategy for repeated serial assessments of tissue vasculature.

When following KS lesions over time, the vascular / metabolic changes in the lesion are important. An additional parameter is of interest when following the treatment over time, which is the structure of the lesion. Optical Coherence Tomography (OCT) is a non-invasive, non-contact optical imaging technology, which provides this desired structural information with high resolution and in three dimensions (3D) over the area of interest. Therefore we aim to combine OCT with multi-spectral data, relating the metabolic state of the tissue with structure. By doing this, we hypothesize that we can not only get deeper understanding of tissue vasculature, but that we can also improve the multi-spectral imaging modality, by using the structure as prior information for the reconstruction.