Imaging tumor vascularity by tracing single microbubbles

Abstract

In high frequency B-Mode images the circulation of single microbubbles through tumor vessels can be observed. We propose the identification of these individual MBs to image tumor vascularity. In addition to the morphological information, a spatial map of the perfusion can be obtained with this technique. The method is tested on B-Mode images of a tumor xenograft in a nude mouse. Circulation of BR38 microbubbles (Bracco, Geneva, Switzerland) was imaged with a Vevo 2100 small animal imaging system at 40 MHz transmit frequency (Visualsonics, Toronto, Canada). Bubbles were identified by filtering subtraction images with a size selective Difference of Gaussians filter. An image of the vascularity is created by mapping the microbubble centroid positions. This map allows a quantitative evaluation of perfused area and the number of microbubbles passing through each pixel. Our first results show that the identification of single bubbles is feasible. The images show improved vessel resolution in comparison to standard maximum intensity persistence images.