Direct echo PIV flow vector mapping on ultrasound DICOM images

Echo Particle Image Velocimetry (Echo PIV) has great potential as a non-invasive, simple-to-use hemodynamic vascular profiling technique. Results generated through the analysis of radio frequency (RF) data have shown to be accurate under in vitro and in vivo testing settings. However, most ultrasound (US) imaging systems do not provide RF data but provide image output using the DICOM standard. An important issue when analyzing DICOM-coded images using the Echo PIV method is how to reduce the effects of noise introduced during post-processing. In this work, a new Echo PIV analysis algorithm was developed to analyze DICOM images. In the algorithm, a quasi-linear transform first operates on the cross-correlation maps of each interrogation window, after which a modified ensemble average method is used to estimate time-averaged velocity field from multiple image pairs. This approach was validated against theory using measurements of fully developed laminar pipe flow, thus indicating that DICOM-image based Echo PIV is possible. We further demonstrate that the instantaneous velocity maps of certain pulsatile flow configurations were actually improved by the new method compared to the RF-based Echo PIV results.