High-Resolution Ultrasound Characterization of Local Scattering in Cancer Tissue

Abstract

Ultrasound (US) has afforded an approach to tissue characterization for more than a decade. The challenge is to reveal hidden patterns in the US data that describe tissue function and pathology that cannot be seen in conventional US images. Our group has developed a high-resolution analysis technique for tissue characterization termed H-scan US, an imaging method used to interpret the relative size of acoustic scatterers. In the present study, the objective was to compare local H-scan US image intensity with registered histologic measurements made directly at the cellular level. Human breast cancer cells (MDA-MB 231, American Type Culture Collection, Manassas, VA, USA) were orthotopically implanted into female mice (N = 5). Tumors were allowed to grow for approximately 4 wk before the study started. In vivo imaging of tumor tissue was performed using a US system (Vantage 256, Verasonics Inc., Kirkland, WA, USA) equipped with a broadband capacitive micromachined ultrasonic linear array transducer (Kolo Medical, San Jose, CA, USA). A 15-MHz center frequency was used for plane wave imaging with five angles for spatial compounding. H-scan US image reconstruction involved use of parallel convolution filters to measure the relative strength of backscattered US signals. Color codes were applied to filter outputs to form the final H-scan US image display. For histologic processing, US imaging cross-sections were carefully marked on the tumor surface, and tumors were excised and sliced along the same plane. By use of optical microscopy, whole tumor tissue sections were scanned and digitized after nuclear staining. US images were interpolated to have the same number of pixels as the histology images and then spatially aligned. Each nucleus from the histologic sections was automatically segmented using custom MATLAB software (The MathWorks Inc., Natick, MA, USA). Nuclear size and spacing from the histology images were then compared with local H-scan US image features. Overall, local H-scan US image intensity exhibited a significant correlation with both cancer cell nuclear size (R² > 0.27, p < 0.001) and the inverse relationship with nuclear spacing (R² > 0.17, p < 0.001).

Keywords: Acoustic scatterers; Cancer; H-Scan ultrasound; Histologic analysis; Tissue characterization; Ultrasound imaging.

Figure 1.

Schematic diagram of the signal processing strategy used to generate and display the H-scan ultrasound (US) images. Backscattered US signals (radiofrequency, RF, format) are processed using a matched filter approach using Gaussian-weighted Hermit polynomial functions of order 2, 6, and 10 (GH₂, GH₆, and GH₁₀, respectively). After envelope detection, filter outputs are assigned to the red (R), green (G), or blue (B) channel of a standard RGB colormap format before final H-scan US image display.

Figure 2.

Schematic diagram of the histological image processing strategy for automated cancer cell nuclear segmentation. Segmentation was performed using adaptive thresholding. To improve segmentation results, a morphological operation was applied before mean nucleus diameter was estimated from the major and minor axis lengths.

Figure 3.

Illustration of the cancer cell nuclear spacing estimation technique based on lines of connectivity between each target nucleus and other surrounding directly connected nuclei. (A) Examples of a nucleus-to-nucleus clear path ratio (N2NPR) is shown for (A) All nuclei of N2NPR > 97%, and a selected nucleus of (B) N2NPR = 100%, (C) N2NPR > 97%, and (D) N2NPR > 37%. Numbers represent the mean spacing (in pixels) between each target nucleus and connected nuclei.

Figure 4.

Matched (A) B-scan US, (B) H-scan US, and (C) histological images. Lower resolution US images were upsampled and then aligned with histology. To offset any registration errors, three different kernel sizes of 160 × 160, 320 × 320, or 800 × 800 μm were used to partition both images into a fully connected set of spaced region-of-interests (ROIs).

Figure 5.

Example (A) B-scan US, (B) H-scan US, and (C) histological images after partitioning into a fully connected set of spaced ROIs that were each 160 × 160 μm in size. Scale bar indicates image intensity.

Figure 6.

(A) Individual scatterplots between local H-scan US image intensity and nuclear size measurements from matched histological subregions. Regression analysis reveals a positive linear relationship (sold line). A total of five subjects and three different-sized kernel sizes of 160 × 160, 320 × 320, or 800 × 800 μm were studied. Subject averages are presented for comparison between registered (B) H-scan US and (C) B-scan US image intensities.

Figure 7.

(A) Individual scatterplots between local H-scan US image intensity and nuclear spacing measurements from matched histological subregions. Regression analysis reveals a negative linear relationship (sold line). Subject averages are presented for comparison between registered (B) H-scan US and (C) B-scan US image intensities.

Figure 8.

Normalized histogram (probability density function, PDF) of (A) H-scan US image intensity, (B) cancer cell nuclear size, and (C) cancer cell spacing measurements and corresponding cumulative distribution function (CDF) for each.