A Pyramidal Optical Flow Method Based on Dense Sift Feature Maps for Motion Correction in Ultrafast Power Doppler Imaging

The visualization and quantitation of microvessels attract increasing attention because microvessels density and hemodynamics are closely related to disease and neural development. Ultrafast power Doppler imaging (uPDI) is a non-invasive technology to image and quantify microvessels with high sensitivity. However, it is sensitive to tissue motion, which is unavoidable during in-vivo imaging. Current mainstream motion correction methods such as normalized cross-correlation (NCC) are based on intensity correlation. However, due to the complex motion of tissue, especially nonrigid motion, the intensity-based correlation is not reliable because the intensity modern around a given pixel may undergo significant changes before and after motion. In addition, NCC-based motion correction algorithms often cost a lot of time. Therefore, in this study, we introduce a pyramidal optical flow method based on dense SIFT feature maps into uPDI to achieve accurate and fast motion correction. SIFT features are robust to rotation, scaling transformation as well as noise, thus SIFT features based correlation has stable discrimination performance when dealing with complex motions. Dense SIFT feature maps are extracted for fixed and moving ultrasound images and an optical flow method is applied to them to estimate the motion. To reduce the computational complexity, a pyramidal structure is employed to estimate the motion from coarser to finer levels. The proposed method is compared with non-rigid NCC and a phase correlation-based motion correction method on a batch of ultrasound images of spinal cord tumors. It achieves the lowest full width at half maxima (FWHM) and can be implemented in several seconds. The results suggest that the proposed method can be a competitive motion correction tool for uPDI.