Gale A. Sisney

Fluorescence Spectroscopy: An Adjunct Diagnostic Tool to Image-Guided Core Needle Biopsy of the Breast

We explored the use of a fiber-optic probe for in vivo fluorescence spectroscopy of breast tissues during percutaneous image-guided breast biopsy. A total of 121 biopsy samples with accompanying histological diagnosis were obtained clinically and investigated in this study. The tissue spectra were analyzed using partial least-squares analysis and represented using a set of principal components (PCs) with dramatically reduced data dimension. For nonmalignant tissue samples, a set of PCs that account for the largest amount of variance in the spectra displayed correlation with the percent tissue composition. For all tissue samples, a set of PCs was identified using a Wilcoxon rank-sum test as showing statistically significant differences between: 1) malignant and fibrous/benign; 2) malignant and adipose; and 3) malignant and nonmalignant breast samples. These PCs were used to distinguish malignant from other nonmalignant tissue types using a binary classification scheme based on both linear and nonlinear support vector machine (SVM) and logistic regression (LR). For the sample set investigated in this study, the SVM classifier provided a cross-validated sensitivity and specificity of up to 81% and 87%, respectively, for discrimination between malignant and fibrous/benign samples, and up to 81% and 81%, respectively, for discriminating between malignant and adipose samples. Classification based on LR was used to generate receiver operator curves with an area under the curve (AUC) of 0.87 for discriminating malignant versus fibrous/benign tissues, and an AUC of 0.84 for discriminating malignant from adipose tissue samples. This study demonstrates the feasibility of performing fluorescence spectroscopy during clinical core needle breast biopsy, and the potential of this technique for identifying breast malignancy in vivo.

Axial-Shear Strain Imaging for Differentiating Benign and Malignant Breast Masses

Axial strain imaging has been utilized for the characterization of breast masses for over a decade; however, another important feature namely the shear strain distribution around breast masses has only recently been used. In this article, we examine the feasibility of utilizing in vivo axial-shear strain imaging for differentiating benign from malignant breast masses. Radio-frequency data was acquired using a VFX 13-5 linear array transducer on 41 patients using a Siemens SONOLINE Antares real-time clinical scanner at the University of Wisconsin Breast Cancer Center. Free-hand palpation using deformations of up to 10% was utilized to generate axial strain and axial-shear strain images using a two-dimensional cross-correlation algorithm from the radio-frequency data loops. Axial-shear strain areas normalized to the lesion size, applied strain and lesion strain contrast was utilized as a feature for differentiating benign from malignant masses. The normalized axial-shear strain area feature estimated on eight patients with malignant tumors and 33 patients with fibroadenomas was utilized to demonstrate its potential for lesion differentiation. Biopsy results were considered the diagnostic standard for comparison. Our results indicate that the normalized axial-shear strain area is significantly larger for malignant tumors compared with benign masses such as fibroadenomas. Axial-shear strain pixel values greater than a specified threshold, including only those with correlation coefficient values greater than 0.75, were overlaid on the corresponding B-mode image to aid in diagnosis. A scatter plot of the normalized area feature demonstrates the feasibility of developing a linear classifier to differentiate benign from malignant masses. The area under the receiver operator characteristic curve utilizing the normalized axial-shear strain area feature was 0.996, demonstrating the potential of this feature to noninvasively differentiate between benign and malignant breast masses.

Utility of 6-month Follow-up Imaging after a Concordant Benign Breast Biopsy Result

PURPOSE To determine the utility of 6-month follow-up imaging after benign concordant image-guided percutaneous breast biopsy results. MATERIALS AND METHODS The institutional review board approved this retrospective, HIPAA-compliant study; informed consent was waived. Findings from consecutive stereotactic and ultrasonographically guided core breast biopsies performed from 2001 to 2005 were analyzed and included lesions with benign pathologic findings without atypia found to be concordant with imaging at a consensus conference. Rebiopsy recommendation rates and positive predictive values (PPVs) for detecting malignancy at each follow-up interval were measured and compared by using a two-tailed Fisher exact test. RESULTS In 2244 biopsies, lesions in 1465 were benign, concordant, and not excised. In 1057 of 1465 (72.2%) biopsies with imaging follow-up (average, 26.4 months; range, 4.0-49.9 months), recommended rebiopsy rates were 0.8% (four of 526), 0.5% (three of 588), and 1.0% (eight of 802) at 6-month, 12-month, and long-term follow-up intervals, respectively. When the initial follow-up did not occur until 12 months, the recommended rebiopsy rate was 0.9% (three of 322), compared with 0.8% (four of 526) at 6 months (P > .99), and no malignancies were found in either group. One malignancy was detected at the long-term follow-up interval (PPV for excision recommended, 12% [one of eight]; PPV for excision performed, 20% [one of five]). CONCLUSION Because rebiopsy recommendation rates and PPVs did not differ in the 6- and 12-month groups, a 6-month follow-up imaging examination, in the context of a formal concordancy consensus conference, may not contribute to improved breast cancer diagnosis. SUPPLEMENTAL MATERIAL http://radiology.rsna.org/lookup/suppl/doi:10.1148/radiol.10091824/-/DC1.

Feasibility of near-infrared diffuse optical spectroscopy on patients undergoing imageguided core-needle biopsy.

We describe a side-firing fiber optic sensor based on near-infrared spectroscopy for guiding core needle biopsy diagnosis of breast cancer. The sensor is composed of three side firing optical fibers (two source fibers and one detection fiber), providing two source-detector separations. The entire assembly is inserted into a core biopsy needle, allowing for sampling to occur at the biopsy site. A multi-wavelength frequency-domain near-infrared instrument is used to collect diffuse reflectance in the breast tissue through an aperture on the biopsy needle before the tissue is removed for histology. Preliminary in vivo measurements performed on 10 normal or benign breast tissues from 5 women undergoing stereo- or ultrasound-guided core needle biopsy show the ability of the system to determine tissue optical properties and constituent concentrations, which are correlated with breast tissue composition derived from histopathology.

Impact of axillary ultrasound and core needle biopsy on the utility of intraoperative frozen section analysis and treatment decision making in women with invasive breast cancer

BACKGROUND Our objective was to evaluate the impact of preoperative axillary ultrasound and core needle biopsy (CNB) on breast cancer treatment decision making. A secondary aim was to evaluate the impact on the utility of intraoperative sentinel lymph node (SLN) frozen section. METHODS A review of 84 patients with clinically negative axilla who underwent axillary ultrasound was performed. Sensitivity, specificity, and positive/negative predictive value for axillary ultrasound with CNB was calculated. RESULTS Thirty-one (37%) had suspicious nodes. Of 27 amenable to CNB, 12 (14%) were malignant, changing treatment plans. The sensitivity of ultrasound and CNB was 54% and specificity 100%; the positive and negative predictive values were 100% and 80%, respectively. In 41 patients with normal ultrasounds who underwent SLN frozen section, 10 (24%) were positive. CONCLUSIONS Preoperative axillary ultrasound impacts treatment decision making in 14%. With a sensitivity of 54%, it is a useful adjunct to, but not replacement for, SLN biopsy. Frozen section remains of utility even after a negative axillary ultrasound.

A novel MR-guided interventional device for 3D circumferential access to breast tissue

MRI is rapidly growing as a tool for image-guided procedures in the breast such as needle localizations, biopsy, and cryotherapy. The ability of MRI to resolve small (<1cm) lesions allows earlier detection and diagnosis than with ultrasound. Most MR-guidance methods perform a two-dimensional compression of the breast that distorts tissue anatomy and limits medial access. This work presents a system for localizing breast lesions with 360° access to breast tissue. A novel system has been developed to perform breast lesion localization using MR guidance that uses a 3D radial coordinate system with four degrees of freedom. The device is combined with a novel breast RF coil for improved signal to noise and rotates 360° around the breast to allow medial, lateral, superior, and inferior access minimizing insertion depth to the target. Coil performance was evaluated using a human volunteer by comparing signal to noise from both the developed breast RF coil and a commercial seven-channel breast coil. The system was tested with a breast-shaped gel phantom containing randomly distributed MR-visible targets. MR-compatible localization needles were used to demonstrate the accuracy and feasibility of the concept for breast biopsy. Localization results were classified based on the relationship between the final needle tip position and the lesion. A 3D bladder concept was also tested using animal tissue to evaluate the devices ability to immobilize deformable breast tissue during a needle insertion. The RF breast coil provided signal to noise values comparable to a seven-channel breast coil. The needle tip was in contact with the targeted lesion in 89% (25∕28) of all the trials and 100% (6∕6) of the trials with targeted lesions >6mm. Target lesions were 3-4mm in diameter for 47% (13∕28), 5-6mm in diameter for 32% (9∕28), and over 6mm in diameter for 21% (6∕28) of the trials, respectively. The 3D bladder concept was shown to immobilize a deformable animal tissue phantom during needle insertion. It is concluded that the MR-guidance system accurately localizes small targets on the order of 3-4mm in a breast phantom with 360° rotational access.

Mammographic breast density and serum phytoestrogen levels.

Some forms of estrogen are associated with breast cancer risk as well as with mammographic density (MD), a strong marker of breast cancer risk. Whether phytoestrogen intake affects breast density, however, remains unclear. We evaluated the association between serum levels of phytoestrogens and MD in postmenopausal women. We enrolled 269 women, ages 55–70 yr, who received a screening mammogram and had no history of postmenopausal hormone use. Subjects completed a survey on diet and factors related to MD and provided a blood sample for analysis of 3 phytoestrogens: genistein, daidzein, and coumestrol. We examined whether mean percent MD was related to serum level of phytoestrogens, adjusting for age and body mass index. Genistein and daidzein levels correlated with self-reported soy consumption. Mean percent MD did not differ across women with different phytoestrogen levels. For example, women with nondetectable genistein levels had mean density of 11.0% [95% confidence intervals (CI) = 9.9–12.4], compared to 10.5% (95% CI = 8.0–13.7) and 11.2% (95% CI = 8.7–14.6) for < and ≥median detectable levels, respectively. In a population with relatively low soy intake, serum phytoestrogens were not associated with mammographic density. Additional studies are needed to determine effects of higher levels, particularly given patterns of increasing phytoestrogen intake.

P3D-2 ROC Analysis of Ultrasound Elasticity Imaging of Breast Abnormalities

Elasticity (mechanical strain) imaging is under rapid development as a tool to improve the specificity of breast ultrasound imaging. Elasticity imaging exploits the fact that benign and malignant breast disease cause inherently different tissue stiffness. We performed a retrospective observer study to determine if elasticity imaging can improve radiologists risk assessment for breast masses over risk assessed with conventional B-mode imaging alone. The best examples of 50 malignant and 48 benign cases were chosen that also represent the distribution of disease from the full data set of 403 breast masses. Other studies precede ours, but our experimental design is unique and this is the first multi-institutional study with multiple expert readers blinded to the tissue pathology. Elasticity imaging was found to increase the ability to assess risk for all observers

TU‐E‐201C‐02: Breast Mass Differentiation Using Axial Shear Strain Imaging

Purpose: Since cancers infiltrate surrounding normal tissue, evoke a desmoplastic, scirrhous reaction and become firmly attached to background tissue, they tend to be less mobile than benign masses like fibroadenomas. We test the feasibility of using in‐vivo axial‐shear strain features to determine if the bonding of masses to background tissue can differentiate benign from malignant. Method and Materials: Radiofrequency data was acquired using a VFX 13‐5 linear array transducer using a Siemens SONOLINE Antares real‐time clinical scanner. Data were acquired for 41 biopsy‐proven breast masses (8 malignant tumors and 33 benign fibroadenomas). Free‐hand palpation using the transducer and deformations of up to 10% was utilized for data acquisition. A two‐dimensional cross‐correlation algorithm was used to generate axial strain and axial‐shear strain images. Axial‐shear strain values normalized to the breast mass dimensions, applied strain and strain contrast were utilized to calculate the feature called the “normalized axial‐shear strain area”, for differentiating benign from malignant masses. Results: The normalized axial‐shear strain area is significantly larger for malignant masses when compared to benign fibroadenomas. Scatter plots of the normalized axial‐shear strain area demonstrates the feasibility for differentiating benign from malignant masses. Receiver operator characteristic analysis demonstrates the improvement in the classification obtained using the normalized axial‐shear strain area. The area under the ROC curve of 0.996 suggests that this feature can effectively differentiate malignant tumors from the benign masses. Conclusions: Axial‐shear strain images may provide important additional information which along with currently utilized axial strain and B‐mode images may improve differentiation of benign and malignant breast masses. This work was supported by Komen grant BCTR0601153 and NTH‐NCI grants R01CA112192‐S103, R01CA100373 and R01 CA111289.

Axial shear strain imaging for breast mass differentiation

Breast cancer remains the second-leading cause of cancer deaths in women, and over 200,000 new cases of invasive breast cancer are expected in the USA this year. Very promising data demonstrate that axial strain imaging has an important role in breast tissue classification However, another important parameter; the shear strain has only recently been recognized as having great potential. We examine the feasibility of utilizing in-vivo axial shear strain imaging for differentiating benign from malignant breast masses. A VFX13-5 linear array transducer was utilized to acquire in-vivo radiofrequency echo data on 41 patients using a Siemens SONOLINE Antares real-time clinical scanner at the University of Wisconsin Breast Center. Free-hand palpation imaging with deformation up to 10% was utilized to acquire radiofrequency data loops to generate strain images. In this study, we report on 8 malignant tumors and 33 fibroadenomas to demonstrate the potential of shear strain imaging, compared to biopsy results that were considered the diagnostic standard. Axial strain and axial component of shear strain are estimated using an algorithm based on 2D cross-correlation. Areas of the axial-shear strain, normalized to the lesion size, applied strain and strain contrast was utilized for differentiating benign from malignant masses. Our results on 40 patients indicate that the normalized axial-shear strain area is significantly larger for malignant tumors when compared to benign fibroadenomas. Axial-shear strain pixel values greater than a specified threshold, including only those with correlation coefficient values greater than 0.75, were overlaid on the corresponding B-mode image to aid in diagnosis. Scatter plot of the normalized area feature demonstrate the feasibility of developing a linear classifier to differentiate benign from malignant masses. The area under the Receiver Operating Characteristic curve using the normalized shear strain area parameter was 0.996.