The physical source of adherent microbubble signal separation in singular spectrum-based targeted molecular imaging (SiSTM) of large vessels

Real-time ultrasound-based molecular imaging in large vessels has the promise for early detection and diagnosis of stroke. Isolation of echo signals from adherent microbubbles (AMBs) from signals attributed to free microbubbles and tissue structures is critical to the success of these imaging techniques. Singular spectrum-based targeted molecular imaging (SiSTM) is a recently proposed method that uses changes in statistical dimensionality, as quantified by the normalized singular spectrum area (NSSA), to more effectively separate components compared with methods that use frequency-domain filtering alone. However, the precise physical mechanism responsible for NSSA-based signal separation of AMBs and vessel wall is still unknown. In this paper, in vitro flow phantom experiments were performed to elucidate the physical mechanism responsible for NSSA-based separation. NSSA and image intensity of AMBs were measured with different radiation forces, microbubble concentrations and flow rates. In the absence of flow, NSSA shifts between vessel wall and AMBs with increasing radiation force and microbubble concentration were less than 2.5% and 10% of typical values (i.e. ~ 0.2) observed from SiSTM, respectively. With flow, and thus with shear forces, the NSSA shift between vessel wall and AMBs was close to 0.2. With different flow rates, the steady-state NSSA level increased with higher flow shear forces. Consequently, our results indicate that the physical source responsible for NSSA-based signal separation is shear forces from flow.