Lacefield, James C.

During contrast-enhanced ultrasound (CEUS) imaging of tumor microvascular perfusion, nearby arteriole enhancement can be a dominant feature of the wash-in kinetics that obscures the perfusion characteristics of the capillary bed. This presentation demonstrates that statistical wash-in curves generated using our compound CEUS speckle model1 exhibit two distinct phases corresponding to “fast-flow” and “slow-flow” enhancement that can be detected by fitting a linear combination of two monoexponential functions with different time constants. CEUS cine loops were acquired from a patient-derived xenograft model of renal cell carcinoma, where fresh tumor fragments were engrafted into the chicken embryo chorioallantoic membrane. Subharmonic CEUS images were acquired at 18 MHz using a destruction-replenishment protocol. Enhancement of manually segmented tumor cross-sections was analyzed offline in MATLAB. The CEUS cine loops from this xenograft model frequently exhibited in-plane arteriole enhancement. The two-phase fit discriminated a slow-flow component in 26.6% of time-intensity wash-in curves and 62.9% of the statistical wash-in curves, leading to a decrease in estimated tumor blood volume by 22.1% and 24.7%, respectively, compared to a simple monoexponential fit. These results suggest that conventional CEUS processing may frequently overestimate capillary blood volume. 1M.R. Lowerison et al., Med. Phys. 44, 99-111 (2017).