Mark B. Williams

Tomographic mammography using a limited number of low-dose cone- beam projection images

A method is described for using a limited number (typically 10-50) of low-dose radiographs to reconstruct the three-dimensional (3D) distribution of x-ray attenuation in the breast. The method uses x-ray cone-beam imaging, an electronic digital detector, and constrained nonlinear iterative computational techniques. Images are reconstructed with high resolution in two dimensions and lower resolution in the third dimension. The 3D distribution of attenuation that is projected into one image in conventional mammography can be separated into many layers (typically 30-80 1-mm-thick layers, depending on breast thickness), increasing the conspicuity of features that are often obscured by overlapping structure in a single-projection view. Schemes that record breast images at nonuniform angular increments, nonuniform image exposure, and nonuniform detector resolution are investigated in order to reduce the total x-ray exposure necessary to obtain diagnostically useful 3D reconstructions, and to improve the quality of the reconstructed images for a given exposure. The total patient radiation dose can be comparable to that used for a standard two-view mammogram. The method is illustrated with images from mastectomy specimens, a phantom, and human volunteers. The results show how image quality is affected by various data-collection protocols.

Noise power spectra of images from digital mammography detectors

Noise characterization through estimation of the noise power spectrum (NPS) is a central component of the evaluation of digital x-ray systems. We begin with a brief review of the fundamentals of NPS theory and measurement, derive explicit expressions for calculation of the one- and two-dimensional (1D and 2D) NPS, and discuss some of the considerations and tradeoffs when these concepts are applied to digital systems. Measurements of the NPS of two detectors for digital mammography are presented to illustrate some of the implications of the choices available. For both systems, two-dimensional noise power spectra obtained over a range of input fluence exhibit pronounced asymmetry between the orthogonal frequency dimensions. The 2D spectra of both systems also demonstrate dominant structures both on and off the primary frequency axes indicative of periodic noise components. Although the two systems share many common noise characteristics, there are significant differences, including markedly different dark-noise magnitudes, differences in NPS shape as a function of both spatial frequency and exposure, and differences in the natures of the residual fixed pattern noise following flat fielding corrections. For low x-ray exposures, quantum noise-limited operation may be possible only at low spatial frequency. Depending on the method of obtaining the 1D NPS (i.e., synthetic slit scanning or slice extraction from the 2D NPS), on-axis periodic structures can be misleadingly smoothed or missed entirely. Our measurements indicate that for these systems, 1D spectra useful for the purpose of detective quantum efficiency calculation may be obtained from thin cuts through the central portion of the calculated 2D NPS. On the other hand, low-frequency spectral values do not converge to an asymptotic value with increasing slit length when 1D spectra are generated using the scanned synthetic slit method. Aliasing can contribute significantly to the digital NPS, especially near the Nyquist frequency. Calculation of the theoretical presampling NPS and explicit inclusion of aliased noise power shows good agreement with measured values.

Analysis of the detective quantum efficiency of a developmental detector for digital mammography.

We are developing a modular detector for applications in full field digital mammography and for diagnostic breast imaging. The detector is based on a design that has been refined over the past decade for applications in x-ray crystallography [Kalata et al., Proc. SPIE 1345, 270-279 (1990); Phillips et al. ibid. 2009, 133-138 (1993), Phillips et al., Nucl. Instrum. Methods Phys. Rev. A 334, 621-630 (1993)]. The full field mammographic detector, currently undergoing clinical evaluation, is formed from a 19 cm x 28 cm phosphor screen, read out by a 2 x 3 array of butted charge-coupled device (CCD) modules. Each 2k x 2k CCD is optically coupled to the phosphor via a fiber optic taper with dimensions of 9.4 cm x 9.4cm at the phosphor. This paper describes the imaging performance of a two-module prototype, built using a similar design. In this paper we use cascaded linear systems analysis to develop a model for calculating the spatial frequency dependent noise power spectrum (NPS) and detective quantum efficiency (DQE) of the detector using the measured modulation transfer function (MTF). We compare results of the calculation with the measured NPS and DQE of the prototype. Calculated and measured DQEs are compared over a range of clinically relevant x-ray exposures and kVps. We find that for x-ray photon energies between 10 and 28 keV, the detector gain ranges between 2.5 and 3.7 CCD electrons per incident x-ray, or approximately 5-8 electrons per absorbed x ray. Using a Mo/Mo beam and acrylic phantom, over a detector entrance exposure range of approximately 10 to 80 mR, the volume under the measured 2-d NPS of the prototype detector is proportional to the x-ray exposure, indicating quantum limited performance. Substantial agreement between the calculated and measured values was obtained for the frequency and exposure dependent NPS and DQE over a range of tube voltage from 25 to 30 kVp.

Optimization of exposure parameters in full field digital mammography

Optimization of exposure parameters (target, filter, and kVp) in digital mammography necessitates maximization of the image signal-to-noise ratio (SNR), while simultaneously minimizing patient dose. The goal of this study is to compare, for each of the major commercially available full field digital mammography (FFDM) systems, the impact of the selection of technique factors on image SNR and radiation dose for a range of breast thickness and tissue types. This phantom study is an update of a previous investigation and includes measurements on recent versions of two of the FFDM systems discussed in that article, as well as on three FFDM systems not available at that time. The five commercial FFDM systems tested, the Senographe 2000D from GE Healthcare, the Mammomat Novation DR from Siemens, the Selenia from Hologic, the Fischer Senoscan, and Fujis 5000MA used with a Lorad M-IV mammography unit, are located at five different university test sites. Performance was assessed using all available x-ray target and filter combinations and nine different phantom types (three compressed thicknesses and three tissue composition types). Each phantom type was also imaged using the automatic exposure control (AEC) of each system to identify the exposure parameters used under automated image acquisition. The figure of merit (FOM) used to compare technique factors is the ratio of the square of the image SNR to the mean glandular dose. The results show that, for a given target/filter combination, in general FOM is a slowly changing function of kVp, with stronger dependence on the choice of target/filter combination. In all cases the FOM was a decreasing function of kVp at the top of the available range of kVp settings, indicating that higher tube voltages would produce no further performance improvement. For a given phantom type, the exposure parameter set resulting in the highest FOM value was system specific, depending on both the set of available target/filter combinations, and on the receptor type. In most cases, the AECs of the FFDM systems successfully identified exposure parameters resulting in FOM values near the maximum ones, however, there were several examples where AEC performance could be improved.

Dual-Modality Breast Tomosynthesis

PURPOSEnTo evaluate the clinical performance of a hybrid scanner that uses dual-modality tomosynthesis (DMT) and technetium 99m sestamibi to provide coregistered anatomic and functional breast images in three dimensions.nnnMATERIALS AND METHODSnA prospective pilot evaluation of the scanner was performed in women scheduled to undergo breast biopsy after institutional review board approval and informed consent were obtained. All subject data were handled in compliance with the rules and regulations concerning the privacy and security of protected health information under HIPAA. The study included 17 women (mean age, 53 years; age range, 44-67 years) and 21 biopsy-sampled lesions. Results of DMT scanning were compared with histopathologic results for the 21 lesions.nnnRESULTSnOf the 21 lesions, seven were malignant, and 14 were benign. Among the 13 subjects with one lesion each, three had positive biopsy results, and 10 had negative biopsy results. Among the four subjects with two lesions, the biopsy results were as follows: bilateral in one, both negative; bilateral in one, both positive; unilateral in two, one positive and one negative. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of DMT scanning were 86%, 100%, 100%, 93%, and 95%, respectively.nnnCONCLUSIONnPilot clinical evaluation of the DMT scanner suggests that it is a feasible and accurate method with which to detect and diagnose breast cancer. Systems such as the DMT scanner that are designed specifically for three-dimensional multimodality breast imaging could make possible some of the advances in tumor detection, localization, and characterization of breast cancer that are now being observed with whole-body three-dimensional hybrid systems, such as positron emission tomography/computed tomography (CT) or single photon emission computed tomography/CT.

Neutrophil targeting heterobivalent SPECT imaging probe: cFLFLF-PEG-TKPPR-99mTc.

A new heterobivalent peptide ligand specifically targeting polymorphonuclear leukocytes (PMNs) with favorable pharmacological parameters to monitor sites of inflammation for imaging is designed. The detailed synthesis, characterization, and pharmacological evaluation of the ligands are reported here. Two separate peptide binding ligands for formyl peptide and tuftsin receptors were chosen to link together based on the high expression levels of the two receptors on activated PMNs The heterobivalency and pegylated links were incorporated in the structural design to improve the sensitivity of the detection and to improve the bioavailability along with blood clearance profile, respectively. Two chemical constructs, cFLFLF-(PEG)(n)-TKPPR-(99m)Tc (n = 4, 12), were evaluated in vitro with human PMNs for binding affinity and bioavailability. As a result, cFLFLF-(PEG)(12)-TKPPR-(99m)Tc was found to have more favorable pharmacological properties and was therefore used for further in vivo studies. Preliminary in vivo assessment of the agent was performed using single gamma emission computed tomography (SPECT) imaging of a mouse model of ear inflammation. The results of these studies indicate cFLFLF-(PEG)(12)-TKPPR-(99m)Tc may be a desirable imaging agent for binding to PMNs to identify sites of inflammation by SPECT.

Intraoperative Imaging Guidance for Sentinel Node Biopsy in Melanoma Using a Mobile Gamma Camera

Objective:To evaluate the sensitivity and clinical utility of intraoperative mobile gamma camera (MGC) imaging in sentinel lymph node biopsy (SLNB) in melanoma. Background:The false-negative rate for SLNB for melanoma is approximately 17%, for which failure to identify the sentinel lymph node (SLN) is a major cause. Intraoperative imaging may aid in detection of SLN near the primary site, in ambiguous locations, and after excision of each SLN. The present pilot study reports outcomes with a prototype MGC designed for rapid intraoperative image acquisition. We hypothesized that intraoperative use of the MGC would be feasible and that sensitivity would be at least 90%. Methods:From April to September 2008, 20 patients underwent Tc99 sulfur colloid lymphoscintigraphy, and SLNB was performed with use of a conventional fixed gamma camera (FGC), and gamma probe followed by intraoperative MGC imaging. Sensitivity was calculated for each detection method. Intraoperative logistical challenges were scored. Cases in which MGC provided clinical benefit were recorded. Results:Sensitivity for detecting SLN basins was 97% for the FGC and 90% for the MGC. A total of 46 SLN were identified: 32 (70%) were identified as distinct hot spots by preoperative FGC imaging, 31 (67%) by preoperative MGC imaging, and 43 (93%) by MGC imaging pre- or intraoperatively. The gamma probe identified 44 (96%) independent of MGC imaging. The MGC provided defined clinical benefit as an addition to standard practice in 5 (25%) of 20 patients. Mean score for MGC logistic feasibility was 2 on a scale of 1–9 (1 = best). Conclusions:Intraoperative MGC imaging provides additional information when standard techniques fail or are ambiguous. Sensitivity is 90% and can be increased. This pilot study has identified ways to improve the usefulness of an MGC for intraoperative imaging, which holds promise for reducing false negatives of SLNB for melanoma.

Combined structural and functional imaging of the breast.

Scintimammography, or single gamma nuclear imaging of the breast, has shown promise as a way of characterizing certain biological properties of suspicious breast masses. Conventional scintimammography, performed using large clinical gamma cameras and prone patient positioning suffers from several drawbacks including poor sensitivity for small (< 1 cm) lesions and no reliable method for correlating scintigraphic findings with those of other imaging modalities. We are developing a system designed to overcome some of these problems. The system combines x-ray mammography with scintimammography on a common gantry. The x-ray and gamma ray images are obtained in quick succession, with the breast in a common configuration under mild compression. A digital x-ray detector is used, permitting rapid assessment of lesion location prior to gamma imaging, and enabling fusion of the x-ray transmission and gamma emission information in a single digital image. In a pilot clinical diagnostic study, the system has demonstrated high pathology-proven accuracy in differentiating benign and malignant masses.

FDG-PET Quantification of Lung Inflammation with Image-Derived Blood Input Function in Mice

Dynamic FDG-PET imaging was used to study inflammation in lungs of mice following administration of a virulent strain of Klebsiella (K.) pneumoniae. Net whole-lung FDG influx constant (K i) was determined in a compartment model using an image-derived blood input function. Methods. K. pneumoniae (~3 x 105 CFU) was intratracheally administered to six mice with 6 other mice serving as controls. Dynamic FDG-PET and X-Ray CT scans were acquired 24 hr after K. pneumoniae administration. The experimental lung time activity curves were fitted to a 3-compartment FDG model to obtain K i. Following imaging, lungs were excised and immunohistochemistry analysis was done to assess the relative presence of neutrophils and macrophages. Results. Mean K i for control and K. pneumoniae infected mice were (5.1 ± 1.2) ×10−3 versus (11.4 ± 2.0) ×10−3u2009min−1, respectively, revealing a 2.24 fold significant increase (P = 0.0003) in the rate of FDG uptake in the infected lung. Immunohistochemistry revealed that cellular lung infiltrate was almost exclusively neutrophils. Parametric K i maps by Patlak analysis revealed heterogeneous inflammatory foci within infected lungs. Conclusion. The kinetics of FDG uptake in the lungs of mice can be noninvasively quantified by PET with a 3-compartment model approach based on an image-derived input function.

Comparison of breast specific gamma imaging and molecular breast tomosynthesis in breast cancer detection: Evaluation in phantoms

PURPOSEnBreast specific gamma imaging or molecular breast imaging (BSGI) obtains 2D images of (99m)Tc sestamibi distribution in the breast. Molecular breast tomosynthesis (MBT) maps the tracer distribution in 3D by acquiring multiple projections over a limited angular range. Here, the authors compare the performance of the two technologies in terms of spatial resolution, lesion contrast, and contrast-to-noise ratio (CNR) in phantom studies under conditions of clinically relevant sestamibi dose and imaging time.nnnMETHODSnThe systems tested were a Dilon 6800 and a MBT prototype developed at the University of Virginia. Both systems comprise a pixelated sodium iodide scintillator, an array of position sensitive photomultipliers, and a parallel hole collimator. The active areas and energy resolution of the systems are similar. System sensitivity, spatial resolution, lesion contrast, and CNR were measured using a Petri dish, a point source phantom, and a breast phantom containing simulated lesions at two depths, respectively. A single BSGI projection was acquired. Five MBT projections were acquired over ±20°. For both modalities, the total scan count density was comparable to that observed for each in typical 10 min human scans following injection of 22 mCi (814 MBq) of (99m)Tc-sestamibi. To assess the impact of reducing the tracer dose, the pixel counts of projection images were later binomially subsampled by a factor of 2 to give images corresponding to an injected activity of approximately 11 mCi (407 MBq). Both unprocessed (pixelated) BSGI projections and interpolated (smoothed) BSGI images displayed by default on the Dilon 6800 workstation were analyzed. Volumetric images were reconstructed from the MBT projections using a maximum likelihood expectation maximization algorithm and extracted slices were analyzed.nnnRESULTSnOver a depth range of 1.5-7.5 cm, BSGI spatial resolution was 5.6-11.5 mm in unprocessed projections and 5.7-12.0 mm in interpolated images. Over the same range, the in-slice MBT spatial resolution was 6.7-9.4 mm. Lesion contrast was significantly improved with MBT relative to BSGI for five out of eight lesions imaged at either the 22 mCi or the 11 mCi dose level (p < 0.05). At both dose levels, significant improvements in CNR with MBT were also found for five out of eight lesions (9.8, 7.8, 6.2 mm lesions at water depth of 1.7 cm and 9.8, 7.8 mm lesions at water depth of 4.5 cm, p < 0.05). The 6.2 and 4.9 mm lesions located at 4.5 cm below the water surface were not visible in either modality at either activity level.nnnCONCLUSIONSnUnder conditions of equal dose, imaging time and similar detectors, compared to BSGI, MBT provided higher lesion contrast, higher CNR, and spatial resolution that was less depth dependent.