Attenuation Estimation and Acoustic Characterization of Mouse Lymph Node Tumor Using High-frequency Ultrasound

Abstract

Purpose: Lymph node (LN) biopsy is the gold standard for diagnosing metastasis. While ultrasound imaging is a non-invasive method for real-time LN metastasis diagnosis and tumor assessment, its accuracy depends on operator skill and system settings. Quantitative ultrasound can characterize tissue microstructure changes due to tumors, offering operator-independent parameters, and one of the quantitative ultrasound methods, the backscatter coefficient, is necessary to compensate for tissue attenuation. However, the change in the attenuation coefficient (AC) in the tumor growth is uncertain. Using in vivo high-frequency ultrasound (25 MHz) measurement and scanning acoustic microscopy (80 and 300 MHz) for ex vivo samples, we aim to investigate how tumor growth is linked to the attenuation and acoustic properties such as acoustic impedance and speed of sound related to ultrasonic wave propagation.

Procedures: FM3 A-Luc mammary carcinoma cells were inoculated into the subiliac LNs of mice, and tumor progression was monitored over time. Bioluminescence imaging was used to assess tumor growth, while ultrasound measurements focused on estimating AC and other acoustic properties.

Results: Results indicated that the mean of AC decreased, and its standard deviation increased as tumors grew, correlating with bioluminescence intensity. Furthermore, acoustic impedance and speed of sound varied between normal and tumor tissues, revealing differences in tissue microstructure from the histopathological images.

Conclusions: The finding of a decrease in AC observed with tumor growth may play a crucial role in enhancing the accuracy of quantitative ultrasound on attenuation compensation, potentially improving the differentiation between metastatic and non-metastatic LNs.

Keywords: Acoustic characterization,; Attenuation coefficient; High-frequency ultrasound; Mouse lymph node; Tumor growth.

Fig. 1.

Schematic image of study design

Fig. 2.

Schematic image of ultrasonic attenuation estimation

Fig. 3.

Spatial distribution of intensity in bioluminescence imaging (groups I, II, and III)

Fig. 4.

AC estimates in days after inoculation of tumor cells. Each marker shows the mean of the local AC in each mouse LN. The linear regression was performed using the mean of the local AC as shown in the dashed line (coefficient of determination = 0.61). The marker shows the mean of the local AC within four groups of experiment. The SD of the local AC was abbreviated for clarity, but the error bar indicates the mean of the SD of the local AC within four groups of experiment. One-way ANOVA and Tukey’s post-hoc test were examined in each experiment; *$p$ < 0.05, day 0 vs. days 14 and 21 in Group I; day 0 vs. day 11 in Group II; day 0 vs. days 13, 16, and 20 in Group III; day 0 vs. days 13 and 16 in Group IV.

Fig. 5.

Scatter plots of the AC estimates and bioluminescence intensity in each mouse LN. Each plot and error bar shows mean and SD of spatial variability within individual data. Tukey's multiple comparison test was used, and Wiseman's rank correlation coefficient $r$ was used for the correlation between AC and luminescence intensity (red line).

Fig. 6.

Example of (a) acoustic impedance, (b) speed of sound, and (c) histopathological images in ex vivo LN on day 21 of group I. The spatial images were aligned by the image registration using the function “registrationEstimator” on MATLAB. Enlarged histopathological images in typical three regions of (c): (d-1) Parenchyma tissue, (d-3) tumor tissue with fibrosis, (d-2) tumor tissue with meristem