Signal processing analysis of twinkling artifact in ultrasound Doppler imaging

The Doppler twinkling artifact (TA) plays a crucial role in detecting some biopsy markers including those made of polymethyl methacrylate (PMMA), facilitating precise lymph node localization in diagnostic and therapeutic procedures. This study presents comprehensive signal processing analysis methods to quantify and characterize the TA phenomenon using raw, per-channel radiofrequency (RF) data acquired from PMMA markers scanned with an L7-4 linear array transducer. We evaluated signals at different stages along the image reconstruction pipeline including (1) beamforming, (2) beamformed RF signal conversion into in-phase/quadrature (IQ) signals, and (3) Doppler velocity calculations and image reconstruction. We analyzed changes of pulses within an ensemble and between ensembles (i.e., frames). We also evaluated the impact of wall filters with various cutoff frequencies on color and power Doppler image reconstruction. Comparative analysis between PMMA markers and a non-twinkling control sample demonstrated distinct TA signatures or lack thereof in color and power Doppler images. The IQ data from PMMA marker locations exhibiting TA signatures showed substantially higher variability compared to non-twinkling samples, with standard deviations approximately 20 times higher between ensembles and 5 times higher within ensembles. This quantitative analysis framework provides valuable insight into the TA signal, potentially leading to more reliable and accurate diagnostic interpretations.