Search

Objective: To detect metastases in freshly excised human lymph nodes (LNs) using three-dimensional (3-D), high-frequency, quantitative ultrasound (QUS) methods, the LN parenchyma (LNP) must be segmented to preclude QUS analysis of data in regions outside the LNP and to compensate ultrasound attenuation effects due to overlying layers of LNP and residual perinodal fat (PNF). Methods: After restoring the saturated radio-frequency signals from PNF using an approach based on smoothing cubic splines, the three regions, i.e., LNP, PNF, and normal saline (NS), in the LN envelope data are segmented using a new, automatic, 3-D, three-phase, statistical transverse-slice-based level-set (STS-LS) method that amends Lankton's method. Due to ultrasound attenuation and focusing effects, the speckle statistics of the envelope data vary with imaged depth. Thus, to mitigate depth-related inhomogeneity effects, the STS-LS method employs gamma probability-density functions to locally model the speckle statistics within consecutive transverse slices. Results: Accurate results were obtained on simulated data. On a representative dataset of 54 LNs acquired from colorectal-cancer patients, the Dice similarity coefficient for LNP, PNF, and NS were 0.938 $\pm$ 0.025, 0.832 $\pm$ 0.086, and 0.968 $\pm$ 0.008, respectively, when compared to expert manual segmentation. Conclusion: The STS-LS outperforms the established methods based on global and local statistics in our datasets and is capable of accurately handling the depth-dependent effects due to attenuation and focusing. Significance: This advance permits the automatic QUS-based cancer detection in the LNs. Furthermore, the STS-LS method could potentially be used in a wide range of ultrasound-imaging applications suffering from depth-dependent effects.

Previous studies have shown that quantitative ultrasound (QUS) methods can provide tissue-microstructure information and are able to successfully detect metastases in human lymph nodes (LNs) harvested from cancer patients. Nevertheless, the gold standard for diagnosis remains pathological evaluation of histology photomicrographs. The goal of the present study is to compare QUS-based and histology-based features which proved to be most valuable for metastatic classification in LNs.

Previous studies by our group have shown that three-dimensional high-frequency quantitative ultrasound methods have the potential to differentiate metastatic lymph nodes from cancer-free lymph nodes dissected from human cancer patients. To successfully perform these methods inside the lymph node parenchyma, an automatic segmentation method is highly desired to exclude the surrounding thin layer of fat from quantitative ultrasound processing and accurately correct for ultrasound attenuation. In high-frequency ultrasound images of lymph nodes, the intensity distribution of lymph node parenchyma and fat varies spatially because of acoustic attenuation and focusing effects. Thus, the intensity contrast between two object regions (e.g., lymph node parenchyma and fat) is also spatially varying. In our previous work, nested graph cut demonstrated its ability to simultaneously segment lymph node parenchyma, fat, and the outer phosphate-buffered saline bath even when some boundaries are lost because of acoustic attenuation and focusing effects. This paper describes a novel approach called graph cut with locally adaptive energy to further deal with spatially varying distributions of lymph node parenchyma and fat caused by inhomogeneous acoustic attenuation. The proposed method achieved Dice similarity coefficients of 0.937+-0.035 when compared to expert manual segmentation on a representative dataset consisting of 115 three-dimensional lymph node images obtained from colorectal cancer patients.

Quantitative acoustic microscopy (QAM) at 500 MHz permits measuring acoustic properties, such as speed of sound (SOS), attenuation (A) and acoustic impedance (Z), of tissue microstructure with a spatial resolution of 4 µm. Although high-frequency QAM has been shown to be a suitable tool for measuring the acoustic properties of several soft tissues, very few studies have been performed at 500 MHz, and none in cancerous lymph nodes (LNs). However, data at such fine resolutions are essential to improve our understanding of ultrasound scattering at lower frequencies (25 MHz) from LNs to detect clinically significant micrometastases. In this study, quantitative acoustic microscopy (QAM) at 500 MHz was performed to obtain 2D maps of speed of sound (c), attenuation (A), acoustic impedance (Z), and other derived acoustical parameters of nodal tissue microstructure with a spatial resolution of 4 µm. Thin section (i.e., 6-µm) of lymph nodes were scanned using a custom-built QAM system based on a novel F-1.08, 500-MHz transducer. The system digitized radio-frequency (RF) signals at 2.5 GHz with 12-bit accuracy.2D QAM maps of the acoustic parameters were obtained using custom a signal-processing algorithm. Following QAM scanning, the samples were stained using hematoxylin and eosin and imaged by light microscopy. The study illustrates that fine-resolution maps of acoustic properties of lymph nodes can be obtained and can provide previously unavailable information. Future studies will investigate the use of 2DZMs to improve the model of ultrasound scattering at 26 MHz. The new lymphnode-specific, ultrasound-scattering models could improve sensitivity and specificity of current QUS approaches for detecting metastatic regions in freshly excised lymph nodes from cancer patients.

Quantitative imaging methods using high-frequency ultrasound (HFU) offer a means of characterizing biological tissue at the microscopic level. Previously, high-frequency, 3-D quantitative ultrasound (QUS) methods were developed to characterize 46 freshly-dissected lymph nodes of colorectal-cancer patients. 3-D ultrasound radiofrequency data were acquired using a 25.6 MHz center-frequency transducer and each node was inked before tissue fixation to recover orientation after sectioning for 3-D histological evaluation. Backscattered echo signals were processed using 3-D cylindrical regions-of-interest (ROIs) to yield four QUS estimates associated with tissue microstructure (i.e., effective scatterer size, acoustic concentration, intercept and slope). These QUS estimates, obtained by parameterizing the backscatter spectrum, showed great potential for cancer detection. In the present study, these QUS methods were applied to 112 lymph nodes from 77 colorectal and gastric cancer patients. Novel QUS methods parameterizing the envelope statistics of the ROIs using Nakagami and homodyned-K distributions were also developed; they yielded four additional QUS estimates. The ability of these eight QUS estimates to classify lymph nodes and detect cancer was evaluated using receiver operating characteristics (ROC) curves. An area under the ROC curve of 0.996 with specificity and sensitivity of 95% were obtained by combining effective scatterer size and one envelope parameter based on the homodyned-K distribution. Therefore, these advanced 3-D QUS methods potentially can be valuable for detecting small metastatic foci in dissected lymph nodes.

Because of time and cost constraints, current histological methods are unable to detect all clinically significant lymph-node metastases efficiently. Therefore, in previous studies, quantitative ultrasound (QUS) methods were developed to detect metastatic regions in freshly excised sentinel lymph nodes from breast-cancer patients using a 26-MHz transducer. A Gaussian form factor was used to obtain QUS estimates for cancer detection. In this study, quantitative acoustic microscopy (QAM) at 250 MHz was performed to obtain 2D maps of speed of sound (c), attenuation (A) and acoustic impedance (Z) of nodal tissue microstructure with a spatial resolution of 7 µm. These maps were used to develop new ultrasound-scattering models to improve understanding of ultrasound scattering in lymph nodes. In total, 17 12-µm thick sections from 12 lymph nodes were scanned using a custom-built QAM system based on an F-1.16, 250-MHz transducer (160-MHz bandwidth). Radio-frequency (RF) signals were digitized at 2.5 GHz with 12-bit accuracy, and 2D QAM maps of Z, A, and c were generated using custom-developed, signal-processing algorithms. Scanned samples were stained using hematoxylin and eosin and imaged by light microscopy. 2D Z maps (2DZMs) were then processed to yield a novel lymph-node-specific scattering model (i.e., form factor), which then was compared to the currently used Gaussian model. The new model was obtained by optimizing an exponential model fit to the Fourier transform of the autocorrelation of the 2DZMs that assumed isotropic scattering. The study illustrates that fine-resolution maps of acoustic properties of lymph nodes can improve the model of ultrasound scattering at 26 MHz. The new lymph-node-specific, ultrasound-scattering model is expected to improve current QUS approaches for detecting metastatic regions in freshly excised sentinel lymph nodes from breast-cancer patients.

To detect metastatic foci in excised human lymph nodes (LNs) using three-dimensional (3D), high-frequency quantitative ultrasound (QUS), the 3D envelope data must be accurately segmented into LN parenchyma (LNP), fat and normal saline (NS). However, automatic segmentation of the 3D data is challenging because of speckle as well as low contrast and intensity inhomogeneities caused by focusing and attenuation effects. We describe a novel method to automatically segment these media using initialization and refinement steps. In the first step, random forest classification (RFC) is employed for initial segmentation of the 3 media. To train the forest classifier and classify each voxel, features including backscattered energy, statistical parameters and contextual information are extracted from LN envelope data. In the second step, the initialization is refined by a 3-phase, local region-based, level-set segmentation method that uses the gamma probability density function as a statistical model of the envelope data. To handle depth-dependent data inhomogeneity efficiently, the gamma distribution parameters are estimated locally in transverse slices. From a database of 54 representative LNs acquired from colorectal-cancer patients, 12 LNs were used to train the random forest classifier, and the 42 remaining LNs were used for evaluation. The Dice similarity coefficient (DSC) was used to compare automatic and manual segmentation. For RFC alone, DSCs were 0.922 ± 0.022, 0.801 ± 0.075 and 0.959 ± 0.013 for LNP, fat and NS, respectively. For initialization and refinement, significantly better DSCs were obtained: 0.937 ± 0.021, 0.824 ± 0.074 and 0.961 ± 0.009 (Wilcoxon signed rank test). Results also demonstrate that accurate QUS estimates can be obtained with automatic segmentation in excised colorectal LNs, thus eliminating the need for operator-dependent, manual segmentation.

Ultrasound (US) data suffer from speckle noise as well as intensity inhomogeneities due to underlying changes in acoustic properties of tissue structure and/or the effects of acoustic focusing and attenuation. This paper describes a 2D and 3D variational level-set method for segmenting such data. To deal with the local statistics of speckle noise, the data term of the level-set energy function is based on local gamma distributions which have shown an ability to model envelope data and gray-level pixel intensities of B-mode clinical images. Local statistics are estimated at a controllable scale using a smooth function with a compact support, a mollifyer, and the method of moments. Compared to manual segmentation, the investigated method provides a high Dice similarity coefficient (DSC) on 3D simulated data, an average DSC of 0.915 on 12 B-mode images of murine tumors acquired with a clinical US system, and average DSCs of 0.920, 0.806 and 0.975 for three media of 54 3D envelope data sets acquired with a high-frequency, focused transducer from dissected human lymph nodes. It also outperforms methods that employ local Gaussian statistics instead of local gamma statistics.

Three-dimensional (3D) quantitative-ultrasound (QUS) methods were recently developed and successfully applied to detect cancerous regions in freshly-dissected lymph nodes (LNs). The 3D high frequency ultrasound (HFU) images obtained from these LNs contain three different parts: LN-parenchyma (LNP), fat, and phosphate-buffered saline (PBS). To apply QUS estimates inside the LNP region, an automatic and accurate algorithm for LNP segmentation is needed. In this paper, we describe a novel, nested-graph-cut (NGC) method that effectively exploits the nested structure of the LN images. To overcome the large variability of the intensity distribution of LNP pixels due to acoustic attenuation and focusing, we further describe an iterative self-updating framework combining NGC and spline-based robust intensity fitting. Dice similarity coefficients of 89.56±8.44% were achieved when the proposed automatic segmentation algorithm was compared to expert manual segmentation on a dataset consisting of 115 LNs.

Scanning acoustic microscopy (SAM) at 250 MHz permits measuring the acoustic properties of tissue microstructure with a spatial resolution of 7 µm. However, only limited data exist for sentinel lymph nodes at these scales. 250-MHz SAM was performed on four fixed and one fresh lymphnode samples. 12-µm sections of deparaffinized, fixed tissue and fresh half nodes were scanned using a custom-built acoustic microscope. 2D SAM maps of speed of sound (c), attenuation (a), and acoustic impedance (Z) were generated using custom signal processing algorithms. Scanned samples then were stained using hematoxylin and eosin and imaged by light microscopy. The spatial resolution and contrast of SAM maps were sufficient to distinguish among tissue regions consisting of lymphocytes, fat cells, and fibrous tissue. Average properties for lymphocyte-rich tissue were c = 1575 ± 57 m/s, a = 8.27 ± 2.45 dB/MHz/cm, and Z = 1.59 ± 0.14 Mrayl. We found a significant increase (p

Quantitative photoacoustics is emerging as a new hybrid modality to investigate diseases and cells in human pathology and cytology studies. Optical absorption of light is the predominant mechanism behind the photoacoustic effect. Therefore, a need exits to characterize the optical properties of specimens and to identify the relevant operating wavelengths for photoacoustic imaging. We have developed a custom low-cost spectrophotometer to measure the optical properties of human axillary lymph nodes dissected for breast-cancer staging. Optical extinction curves of positive and negative nodes were determined in the spectral range of 400 to 1000 nm. We have developed a model to estimate tissue optical properties, taking into account the role of fat and saline. Our results enabled us to select the optimal optical wavelengths for maximizing the imaging contrast between metastatic and noncancerous tissue in axillary lymph nodes.

Abdominal ultrasound is one of the most commonly used imaging methods currently. It is different from other imaging such as computed tomography because surgeons (especially abdominal surgeons, hepatobiliary pancreatic surgeons, surgical oncologists, and minimally invasive surgeons) can and should perform ultrasound examination by themselves.

This work investigates the statistics of the envelope of three-dimensional (3D) high-frequency ultrasound (HFU) data acquired from dissected human lymph nodes (LNs). Nine distributions were employed, and their parameters were estimated using the method of moments. The Kolmogorov Smirnov (KS) metric was used to quantitatively compare the fit of each candidate distribution to the experimental envelope distribution. The study indicates that the generalized gamma distribution best models the statistics of the envelope data of the three media encountered: LN parenchyma, fat and phosphate-buffered saline (PBS). Furthermore, the envelope statistics of the LN parenchyma satisfy the pre-Rayleigh condition. In terms of high fitting accuracy and computationally efficient parameter estimation, the gamma distribution is the best choice to model the envelope statistics of LN parenchyma, while, the Weibull distribution is the best choice to model the envelope statistics of fat and PBS. These results will contribute to the development of more-accurate and automatic 3D segmentation of LNs for ultrasonic detection of clinically significant LN metastases.

Proper staging and treatment of cancer require accurate detection of lymph-node metastases, but current histological methods fail to detect small, but clinically significant metastases. We used novel 3D quantitative ultrasound (QUS) methods to identify metastatic regions in freshly excised lymph nodes from cancer patients. Individual lymph nodes were scanned in 3D using a 26-MHz, single-element, F2 transducer with a 12-mm focal length. QUS methods quantified the backscatter coefficient to yield four estimates in cylindrical regions of interest (ROIs) having equal lengths and diameters ranging from 0.4 to 1 mm. To optimize the tradeoff between QUS-estimate quality and the spatial resolution of the estimates, the effect of ROI size on estimate bias and variance was investigated using a database of 101 lymph nodes of colorectal-cancer patients. Estimates were combined using linear-discriminant approaches and ROC curves were computed to assess classification performance. A Bayesian approach was used to convert the discriminant scores to 3D cancer-probability estimates throughout each lymph node. Analysis indicated that ROIs with a 0.8-mm length and diameter improved spatial resolution and minimally degraded estimate quality with an average variance increase of

Quantitative ultrasound (QUS) methods using high-frequency ultrasound offer a means of investigating biological tissue at the microscopic level. This chapter describes high-frequency, three-dimensional (3D) QUS methods to characterize freshly dissected lymph nodes of cancer patients. 3D ultrasound radio-frequency data were acquired from lymph nodes using a 25.6-MHz center-frequency transducer. Each node was inked prior to tissue fixation to recover orientation after sectioning for 3D histological evaluation. Backscattered echo signals were processed using 3D cylindrical regions-of-interest to yield four QUS estimates associated with tissue microstructure (i.e., effective scatterer size, acoustic concentration, spectral intercept, and spectral slope). QUS estimates were computed following established methods using two scattering models. Then, the four QUS estimates were combined using linear-discriminant analysis to increase classification performance. Finally, the discriminant scores were used to compute a posteriori cancer probability. In this study, more than 400 lymph nodes acquired from more than 250 patients diagnosed with colon, breast, or gastric cancer were processed. Results indicated that metastatic and cancer-free lymph nodes of colon- and gastric-cancer patients could be well classified using these methods and that metastatic regions could potentially be detected and used to guide pathologists towards suspicious regions.

Purpose Detection of metastases in lymph nodes (LNs) is critical for cancer management. Conventional histological methods may miss metastatic foci. To date, no practical means of evaluating the entire LN volume exists. The aim of this study was to develop fast, reliable, operator-independent, high-frequency, quantitative ultrasound (QUS) methods for evaluating LNs over their entire volume to effectively detect LN metastases. Methods We scanned freshly excised LNs at 26 MHz and digitally acquired echo-signal data over the entire three-dimensional (3D) volume. A total of 146 LNs of colorectal, 26 LNs of gastric, and 118 LNs of breast cancer patients were enrolled. We step-sectioned LNs at 50-μm intervals and later compared them with 13 QUS estimates associated with tissue microstructure. Linear-discriminant analysis classified LNs as metastatic or nonmetastatic, and we computed areas (Az) under receiver-operator characteristic curves to assess classification performance. The QUS estimates and cancer probability values derived from discriminant analysis were depicted in 3D images for comparison with 3D histology. Results Of 146 LNs of colorectal cancer patients, 23 were metastatic; Az = 0.952 ± 0.021 (95% confidence interval [CI]: 0.911–0.993); sensitivity = 91.3% (specificity = 87.0%); and sensitivity = 100% (specificity = 67.5%). Of 26 LNs of gastric cancer patients, five were metastatic; Az = 0.962 ± 0.039 (95% CI: 0.807–1.000); sensitivity = 100% (specificity = 95.3%). A total of 17 of 118 LNs of breast cancer patients were metastatic; Az = 0.833 ± 0.047 (95% CI: 0.741–0.926); sensitivity = 88.2% (specificity = 62.5%); sensitivity = 100% (specificity = 50.5%). 3D cancer probability images showed good correlation with 3D histology. Conclusions These results suggest that operator- and system-independent QUS methods allow reliable entire-volume LN evaluation for detecting metastases. 3D cancer probability images can help pathologists identify metastatic foci that could be missed using conventional methods.

Histopathological detection of metastases in dissected lymph nodes of cancer patients is critical to proper staging and management but current histological methods fail to detect small, but clinically-relevant, metastases. Quantitative ultrasound (QUS) permits characterization of tissue microstructure using system-independent estimates. In this study, more than 250 freshly-excised lymph nodes were evaluated using specifically designed QUS methods that demonstrated an ability to guide pathologists towards suspicious regions using an interactive and easy-to-use GUI called Lymph Explorer. Radio-frequency (RF) data were acquired in 3D using a 26-MHz transducer and RF data were processed to yield 13 QUS estimates associated with tissue microstructure. The 13 QUS estimates were combined using a linear discriminant classifier to derive cancer-probability estimates and classification performance was assessed using ROC methods. For gastrointestinal nodes, the areas under the ROC curves (AUCs) exceeded 0.95. Slightly poorer results (AUCs 0.85) were obtained for nodes of breast-cancer patients. Lymph Explorer can interactively display any three orthogonal cross-sectional B-mode images with overlaid color-coded cancer probabilities. In particular, Lymph Explorer permitted localization of small metastases in some partially-metastatic cases. The QUS approach integrated with Lymph Explorer potentially could drastically reduce the current rate of false-negative determinations by efficiently guiding pathologists to suspicious regions in dissected lymph nodes.

Our group is developing a method based on 3D high-frequency ultrasound (HFU) and 3D quantitative ultrasound (QUS) to help pathologists detect micrometastases in freshly-excised lymph nodes of patients with histologically-proven primary cancer. From a signal and image processing perspective, we report on our efforts to acquire and classify lymph-node tissue based on 3D QUS parameter estimates. We evaluated classifier performance against gold-standard histology. Using our database of 134 abdominal cancer-free nodes and 26 fully cancerous abdominal nodes, a conservative threshold gave a sensitivity and specificity of 99.3% and 73.1%, respectively. We also constructed a 3D cancer-likelihood map of a partially metastatic lymph node and compared that map with histology. This representation potentially can be useful for guiding pathologists to suspicious regions requiring histological evaluation.

The purpose of this study was to accurately detect lymph-node micrometastases, i.e., metastatic cancer foci that have a size between 2.0 and 0.2 mm, in nodes excised from colorectal cancer (CRC) patients, and to determine how frequently micrometastases might be missed when standard histological examination procedures are used. A total of 311 lymph nodes were removed and examined from 90 patients with Stage I to IV CRC. The number of slices of histology sections ranged from 6 to 75 per node (average = 25.5; SD = 11.1), which provided a total of 7,943 slices. Lymph nodes were examined in their entire volume at every 50-μm and 100-μm intervals for nodes smaller and larger than 5 mm respectively. The total number of thin sections examined in each node and the number of thin sections where metastatic foci were present were counted. The number of thin sections with metastatic foci and the total number of slices was determined for each node. In addition, the presence or absence of metastatic foci in the “central” slice was determined. Micrometastases were found in 12/311 (3.9%) of all lymph nodes. In the 12 lymph nodes with micrometastases, the rate of metastatic slices over all slices was 39.4% (range = 6.3 to 81.3%; SD = 25.8%) In the central slice of each node, micrometastases were present only in 6 of 12 lymph nodes (50%); accordingly, they were not present in the central slice for half the micrometastatic nodes. These 6 nodes represented 1.9% of the 311 nodes and 11.1% of the 54 metastatic nodes. This study suggests that a significant fraction of micrometastases can be missed by traditional singleslice sectioning; half of the micrometastases would have been overlooked in our data set of 311 nodes.

High-Frequency Quantitative Ultrasound (HFQUS) imaging methods are under investigation to evaluate their ability to detect small metastases ( 5 mm, were photographed. Then a fully automatic method to assemble and orient a 3D histologic volume from a set of 2D images was developed and applied. Identification of the histology sections on each slide relies on a parametric shape modeling of the histologic sections with ellipses. Then a set of rigid transformations were estimated and applied to construct volumetric histologic data. The method was visually evaluated on a set of 50 lymph nodes and is valuable for comparing histologic data to HFQUS estimates in 3D.

Quantitative imaging methods using high-frequency ultrasound (HFU, >20 MHz) offer a means of characterizing biological tissue at the microscopic level. In this study, three-dimensional (3D) quantitative-ultrasound (QUS) methods were developed to detect metastases in freshly-dissected lymph nodes of cancer patients. 3D ultrasound radio-frequency data were acquired using a 26-MHz center-frequency transducer, and each node was inked prior to tissue fixation to recover orientation after sectioning for 3D histological evaluation. Backscattered echo signals were processed using 3D cylindrical regions-of-interest (ROIs) to yield eight QUS estimates associated with tissue microstructure. The first four QUS estimates (i.e., effective scatterer size, acoustic concentration, intercept, and slope) were estimated from a normalized backscattered spectrum, and the other four QUS estimates were obtained by parameterizing the envelope statistics of the ROIs using Nakagami and homodyned-K models. These QUS methods were applied to 145 lymph nodes from 95 colorectal and gastric cancer patients. The ability of these eight QUS estimates to classify lymph nodes and detect cancer was evaluated using ROC curves. An area under the ROC curve of 0.971 with specificity and sensitivity of 91% (using a leave-one-out procedure) were obtained by combining effective scatterer size and one envelope parameter based on the homodyned-K distribution. Therefore, these advanced 3D QUS methods potentially can be valuable for detecting small metastatic foci in dissected lymph nodes.

A reliable means of detecting metastases in lymph nodes is essential for accurate staging of cancer and effective planning of therapy. The standard means of detecting metastases in nodes is histopathologic evaluation of surgically dissected nodes; however, because examinations tend to be limited to thin sections derived from the center of the node, this time‐consuming procedure appears to have high false‐negative rates for metastases that are 2 mm or smaller. We have been investigating high‐frequency (HF) quantitative ultrasound (QUS) methods as a more‐effective means of providing a rapid and reliable detection of metastases in dissected nodes based on their ultrasound‐scattering properties. We acquired HF ultrasound and histological data from dissected lymph nodes of patients with colorectal, gastric, breast, and other cancers. Freshly dissected nodes were scanned in a saline bath using a raster pattern to acquire 3D RF echo‐signal data. Scans utilized a broadband, F‐2, 25.6‐MHz, single‐element transducer with scan vectors separated by 25 μm in X and Y directions. Scanned nodes were color inked to provide references for subsequent orientation, then fixed and serially‐sectioned in their entire volume at 50‐μm intervals. The presence of metastatic foci was determined histologically in every section; evaluation of node status in the center section provided a basis for comparison with conventional methods. To date, we have analyzed the echo signals of more than 240 nodes including abdominal nodes of colorectal and gastric cancer patients and axillary nodes of breast‐cancer patients. 3D images generated from RF data were segmented semi‐automatically to select nodal tissue for analysis. Echo signals from nodal tissue were processed to yield QUS estimates, which included spectral parameters, scatterer‐property estimates, and B‐mode envelope‐signal statistical features. Different histological node architectures were observed and different QUS results were obtained for abdominal compared to axillary nodes. Linear discriminant analysis and ROC‐curve methods were applied to assess the ability of QUS estimates to distinguish cancerous from non‐cancerous nodes. Classification performance was assessed for individual estimates and various linear combinations of estimates. ROC results for axillary as well as abdominal nodes showed excellent classification. For abdominal nodes, the best area under the ROC curve exceeded 0.95. For axillary nodes, the best area approached 0.90. Images based on QUS parameters showed an excellent ability to depict metastatic foci. Results to date strongly suggest that HF QUS methods may provide a clinically valuable means of detecting small metastatic cancers in dissected lymph nodes and significantly improving the sensitivity for metastases that might not be detected using current, standard histopathology procedures. The ability of HF QUS to reveal otherwise missed metastases will enable pathologists to more‐ efficiently focus histological effort on regions of nodes that have a high suspicion of containing cancer and minimizing wasted time evaluating actually cancer‐free nodes. Future studies will investigate the applicability of these methods to detection of nodal metastases in situ.

Histological assessment of lymph nodes excised from cancer patients suffers from an unsatisfactory rate of false-negative determinations. We are evaluating high-frequency quantitative ultrasound (QUS) to detect metastatic regions in lymph nodes freshly excised from cancer patients. Three-dimensional (3D) RF data were acquired from 289 lymph nodes of 82 colorectal-, 15 gastric-, and 70 breast-cancer patients with a custom scanner using a 26-MHz, single-element transducer. Following data acquisition, individual nodes underwent step-sectioning at 50-µm to assure that no clinically significant cancer foci were missed. RF datasets were analyzed using 3D regions-of-interest that were processed to yield 13 QUS estimates including spectral-based and envelope-statistics-based parameters. QUS estimates are associated with tissue microstructure and are hypothesized to provide contrast between non-cancerous and cancerous regions. Leave-one-out classifications, ROC curves, and areas under the ROC (AUC) were used to compare the performance of support vector machines (SVMs) and step-wise linear discriminant analyses (LDA). Results showed that SVM performance (AUC = 0.87) was superior to LDA performance (AUC = 0.78). These results suggest that QUS methods may provide an effective tool to guide pathologists towards suspicious regions and also indicate that classification accuracy can be improved using sophisticated and robust classification tools. [Supported in part by NIH grant CA100183.]

We employed quantitative acoustic microscopy (QAM) to measure acoustic properties of tissue microstructure. 32 QAM datasets were acquired from 2, fresh and 11, deparaffinized, 12-µm-thick lymph-node samples obtained from cancer patients. Our custom-built acoustic microscope was equipped with an F-1.16, 250-MHz transducer having a 160-MHz bandwidth to acquire reflected signals from the tissue and a substrate that intimately contacted the tissue. QAM images with a spatial resolution of 7 µm were generated of attenuation (A), speed of sound (SOS), and acoustic impedance (Z). Samples then were stained using hematoxylin and eosin, imaged by light microscopy, and co-registered to QAM images. The spatial resolution and contrast of QAM images were sufficient to distinguish tissue regions consisting of lymphocytes, fat cells and fibrous tissue. Average properties for lymphocyte-dominated tissue were 1552.6 ± 30 m/s for SOS, 9.53 ± 3.6 dB/MHz/cm for A, and 1.58 ± 0.08 Mrayl for Z. Values for Z obtained from fresh samples agreed well with those obtained from 12-µm sections from the same node. Such 2D images provide a basis for developing improved ultrasound-scattering models underlying quantitative ultrasound methods currently used to detect cancerous regions within lymph nodes. [NIH Grant R21EB016117.]

Histology performed to assess lymph nodes excised during node-dissection surgeries from cancer patients suffers an unsatisfactory rate of false-negative determinations due to labor and time constraints. In this study, more than 300 lymph nodes were scanned in 3D using a 26-MHz high-frequency ultrasound transducer. Following scanning, individual nodes underwent a special histology procedure that involved step-sectioning each node at 50-µm intervals to guarantee that no significant cancer foci were missed. The 3D radio-frequency ultrasound dataset was analyzed using overlapping 3D regions-of-interests that were individually processed to yield thirteen quantitative ultrasound (QUS) estimates associated with tissue microstructure and were hypothesized to show contrast between normal and cancerous regions in lymph nodes. Step-wise linear discriminant analyses were performed to yield an optimal QUS-based classifier. ROC curves and areas under the ROC curves (AUCs) were obtained to assess cancer-detection performance. The AUC for the linear combination of four QUS estimates was 0.83 for a dataset of 110 axillary nodes of breast-cancer patients. Similarly, using five QUS estimates, an AUC of 0.97 was obtained for a dataset of 180 nodes of gastrointestinal-cancer patients. These studies demonstrate that QUS methods may provide an effective tool to guide pathologist towards suspicious regions in lymph nodes.

Human lymph nodes excised from cancer patients during lymphadenectomy can contain small clinically-important metastatic regions that can be missed because conventional histopathology methods do not allow nodes to be examined over their entire volume. In this study, more than 250 lymph nodes were scanned in 3D using a 26-MHz ultrasound transducer before histology processing. Acquired radio-frequency data were processed using 3D regions-of-interest to yield thirteen quantitative ultrasound (QUS) estimates. The QUS estimates are related to tissue microstructure and are hypothesized to be different in normal nodal tissue and metastatic tissue. Four QUS estimates were obtained from backscattered spectra and the remaining nine were derived from envelope statistics. Following ultrasound scanning, serial-section histology was performed at 50-μm intervals to depict cancer foci in 3D. Classification based on QUS estimates was performed using linear-discriminant analyses in a step-wise approach, and areas under ROC ...

RESULTS: Discriminant analysis showed that optimal classification required only 2 to 4 parameters (scatterer size and others). 26/ 160 GI-cancer nodes were metastatic; classification gave an area under the ROC curve of 0.974 0.011 (95% CI: 0.952 – 0.997); sensitivity 99.3% (corresponding specificity 73.1%); sensitivity 100% (corresponding specificity 62.1%). 15/94 breast-cancer nodes were metastatic; classification gave an area under the ROC curve of 0.889 0.046 (95% CI: 0.799 – 0.979); sensitivity 93.3% (corresponding specificity 67.1%); sensitivity 100% (corresponding specificity 41.8%). CONCLUSIONS: Theseresultssuggestthatthisoperator-independent QUS method will allow accurate entire-volume LN examination. 3D reconstruction with QUS parameter images can help pathologists identify metastatic foci that may be missed using conventional histological methods.

High‐frequency quantitative ultrasound (QUS) may offer a reliable means of identifying tumor foci in dissected lymph nodes. Detection of metastases is essential for staging and treatment planning. Conventional histopathology methods do not allow nodes to be examined over their entire volume. Therefore, our objective is to develop QUS methods to improve detection of clinically significant lymph‐node metastases. A single‐element 26‐MHz ultrasound transducer was used to scan and digitally acquire rf echo‐signal data in three‐dimensional (3D) from more than 200 lymph nodes. Thirteen QUS estimates based on backscatter spectra and envelope statistics were computed in 3D. Serial‐sectioning histology was performed at 50‐μm intervals to depict cancer foci in 3D. Classification based on QUS estimates was performed using linear‐discriminant analyzes; areas under ROC curves (AUCs) were computed. The most‐significant QUS estimates for metastases detection were identified. Comparison of the 3D QUS results and 3D histol...