Quantitative Multifrequency Ultrasound Imaging Using Narrowband Pulsing for Tissue Characterization.

Abstract

Objective: Multifrequency ultrasound (MFUS) is a novel quantitative imaging technique that captures backscattered data after transmission of a sequence of frequency-shifted narrowband ultrasound (US) pulses. The goal of this study was to implement a MFUS acquisition sequence and then validate quantitative measurements using an animal model of metabolic dysfunction-associated steatotic liver disease (MASLD).

Materials and methods: US imaging was performed using a programmable research scanner (Vantage NXT 256, Verasonics Inc) equipped with a L11-5v linear array. Radiofrequency (RF) data were acquired from 24 different frequency-shifted narrowband transmit pulses using plane wave imaging with spatial angular compounding. The MFUS gradient was defined as the linear slope between logarithmic B-scan intensity and transmit frequency. The US attenuation coefficient and relative scatterer size (RSS) were derived from the MFUS gradient parameter. A bilayer tissue-mimicking phantom with regions of known US attenuation values of -0.70 and -0.95 dB/cm/MHz, respectively, was used for in vitro MFUS imaging and attenuation coefficient estimation. Custom homogeneous phantoms containing 38 to 89 µm-sized US scatterers were also used to compare RSS estimates. A heterogeneous phantom containing a circular target and background material embedded with 64 and 38 µm-sized scatterers, respectively, was used to measure the contrast-to-noise ratio (CNR) between the two regions for B-scan and MFUS-derived RSS estimates. Preclinical US imaging was performed on mice fed standard chow or a methionine and choline-deficient (MCD) diet to induce liver steatosis ( n =5 animals per group). Mice underwent in vivo MFUS imaging, and then attenuation coefficient and RSS estimates were derived for each mouse liver.

Results: In vitro attenuation estimates were found to be -0.71 ± 0.04 and -0.93 ± 0.03 dB/cm/MHz for a phantom with true attenuation values of -0.70 and -0.95 dB/cm/MHz, respectively. A difference was found in RSS values from homogenous phantoms embedded with different-sized US scatterers and CNR measurements from the heterogeneous inclusion phantom ( P < 0.001). Average CNR values increased from 0.11 to 0.30 when using B-scan and MFUS imaging, respectively. Relative to control animals, in vivo liver imaging studies revealed increased attenuation coefficient and decreased RSS values in mice fed the MCD diet ( P < 0.001). After US imaging, all animals were humanely euthanized and livers excised for histologic analysis.

Conclusion: MFUS imaging of phantom materials revealed a sensitivity to changes in attenuation coefficient and US scatterer size conditions. Preliminary preclinical results validated the performance of MFUS-derived parameters for liver tissue characterization. Changes in attenuation coefficient and RSS values were consistent with previous studies. Overall, MFUS is an innovative imaging technique, and future studies will explore further optimization and potential applications.