A Statistical Model for the Quantification of Microbubbles in Destructive Imaging

Rationale and Objectives: Quantification of targeted ultrasound contrast agents allows for the monitoring of endothelial marker expressions on a molecular level. In this study, a statistical correction is provided, which allows for improved precision in estimating the concentration of microbubbles (MBs) from Doppler images. Doppler imaging can be used to display the destruction of single MBs. However, concentrations will generally be too high to distinguish the individual events resulting in an inaccurate microbubble (MB) quantification. Therefore, a mathematical description of destruction events in Doppler images is developed which yields a correction formula for the concentration estimate from the color pixel density. Methods: The mathematical model is experimentally verified in gelatin phantoms using a high resolution imaging system (Vevo 770) and experimental cyanoacrylate MBs. Sensitive Particle Acoustic Quantification (SPAQ) is used to quantify MB in a defined volume. The SPAQ step size is varied from 32 to 127 μm to demonstrate the validity of the model for high color pixel densities. Results: The corrected acoustic quantification shows the expected linear dependence on the step size and thus the amount of MBs in the images (R2 = 0.95). At SPAQ step sizes up to 127 μm, a MB concentration of 2.7 × 106 MBs/mL can be quantified. Conclusions: The results demonstrate the validity of the proposed correction. Quantification results of the SPAQ technique were considerably improved. The resulting formula is readily applied to SPAQ measurements at no additional expense.