Three-dimensional high-frequency characterization of excised human lymph nodes

Abstract

High-frequency ultrasound (HFU, >20 MHz) offers a means of investigating biological tissue at the microscopic level because the short wavelengths and small focal-zone beam diameters of HFU transducers provide fine-resolution images (<100 ¿m). In this study, three-dimensional (3D) quantitative-ultrasound (QUS) methods were developed and evaluated to detect metastases in freshly-dissected lymph nodes of cancer patients. Detection of metastases is critically important for staging and treatment planning. 3D radio-frequency (RF) data were acquired from scanning dissected lymph nodes using a 26-MHz single-element transducer. For each lymph node, overlapping cylindrical regions-of-interest (ROIs, 1-mm diameter, 1-mm deep) were processed to yield four QUS estimates. Spectral intercept (I) and spectral slope (S) were computed using a straight-line model and effective scatterer size (D) and acoustic concentration (CQ ² ) were estimated using a Gaussian scattering model. 3D QUS images were generated by expressing QUS estimates as color-encoded pixels and overlaying them on conventional 3D B-mode images. QUS estimates were obtained for 82 lymph nodes (65 devoid of metastases and 17 entirely filled by metastases) from 46 patients diagnosed with colon or gastric cancer and a statistical difference (p <0.05) was found for each QUS estimate for metastatic and non-metastatic nodes. Furthermore, areas under the ROC were greater than 0.99 for D or S alone. These initial results suggest that these methods may provide a clinically important means of identifying small metastatic foci that might not be detected using standard pathology procedures.