Anita Nosratieh

Anatomical complexity in breast parenchyma and its implications for optimal breast imaging strategies

PURPOSE The purpose of this investigation was to assess the anatomical noise in breast images using a mathematically derived parameter β as a surrogate for detection performance, across the same patient cohort but in different imaging modalities including mammography, tomosynthesis, and breast CT. METHODS Women who were scheduled for breast biopsy were approached for participation in this IRB and HIPPA-compliant investigation. A total of 23 women had all views of each modality and represent the cohort studied in this investigation. Image data sets across all modalities were analyzed using 1000 regions of interest per image data set, and the anatomical noise power spectrum, NPS(a)(f), was computed and averaged for each breast image data set. After windowing the total noise power spectrum NPS(t)(f) to a specific frequency range corresponding to anatomical noise, the power-law slope (β) of the NPS(a)(f) was computed where NPS(a)(f) = α f(-) (β). RESULTS The value of β was determined for breast CT data sets, and they were 1.75 (0.424), 1.83 (0.352), and 1.79 (0.397), for the coronal, sagittal, and axial views, respectively. For tomosynthesis, β was 3.06 (0.361) and 3.10 (0.315) for the craniocaudal (CC) and medial lateral oblique (MLO) views, respectively. For mammography, these values were 3.17 (0.226) and 3.30 (0.236), for the CC and MLO views, respectively. The values of β for breast CT were significantly different than those for tomosynthesis and mammography (p < 0.001, all 12 comparisons). CONCLUSIONS Based on the parameter β which is thought to describe anatomical noise in breast images, breast CT was shown to have a statistically significant lower β than mammography or tomosynthesis. It has been suggested in the literature that a lower β may correspond to increased cancer detection performance; however, this has yet to be demonstrated unequivocally.

Radiation dosimetry in digital breast tomosynthesis: Report of AAPM Tomosynthesis Subcommittee Task Group 223

The radiation dose involved in any medical imaging modality that uses ionizing radiation needs to be well understood by the medical physics and clinical community. This is especially true of screening modalities. Digital breast tomosynthesis (DBT) has recently been introduced into the clinic and is being used for screening for breast cancer in the general population. Therefore, it is important that the medical physics community have the required information to be able to understand, estimate, and communicate the radiation dose levels involved in breast tomosynthesis imaging. For this purpose, the American Association of Physicists in Medicine Task Group 223 on Dosimetry in Tomosynthesis Imaging has prepared this report that discusses dosimetry in breast imaging in general, and describes a methodology and provides the data necessary to estimate mean breast glandular dose from a tomosynthesis acquisition. In an effort to maximize familiarity with the procedures and data provided in this Report, the methodology to perform the dose estimation in DBT is based as much as possible on that used in mammography dose estimation.

Comprehensive assessment of the slice sensitivity profiles in breast tomosynthesis and breast CT

PURPOSE This study experimentally evaluated the slice sensitivity profile (SSP) and its relationship between acquisition angle, object size, and cone angle. The sensitivity profile metric was used to characterize a breast tomosynthesis systems resolution in the z-axis. The SSP was also measured on a prototype breast computed tomography (bCT) system. METHODS The SSP was measured using brass disks placed within adipose tissue-equivalent breast phantoms. The digital tomosynthesis system (Selenia Dimensions, Hologic Corporation, Bedford, MA) acquires projection images over a 15° angular range and the bCT scanner acquires projection images over a 360° angular range. Angular ranges between 15° and 360° were studied by using a subset of the projection images acquired on the bCT scanner. The SSP was determined by measuring a background-corrected mean gray scale value as a function of the z-position (axis normal to the plane of the detector). RESULTS The results show that SSP improves when the angular acquisition range is increased and the SSP approaches a delta function for angles greater than 180°. Smaller objects have a narrower SSP and the SSP is not significantly dependent on the cone angle. For a 2.5, 5, 10 mm disk, the full width at half maximum of the SSP was 35, 61, 115 mm, respectively, on the tomosynthesis system (at 15°) and was 0.5 mm for all disk diameters on the bCT scanner (at 360°). CONCLUSIONS The SSP is dependent on object size and angular acquisition range. These dependencies are overcome once the angular acquisition range is increased beyond 180°.

Mean glandular dose coefficients (D(g)N) for x-ray spectra used in contemporary breast imaging systems.

To develop tables of normalized glandular dose coefficients D(g)N for a range of anode-filter combinations and tube voltages used in contemporary breast imaging systems. Previously published mono-energetic D(g)N values were used with various spectra to mathematically compute D(g)N coefficients. The tungsten anode spectra from TASMICS were used; molybdenum and rhodium anode-spectra were generated using MCNPX Monte Carlo code. The spectra were filtered with various thicknesses of Al, Rh, Mo or Cu. An initial half value layer (HVL) calculation was made using the anode and filter material. A range of the HVL values was produced with the addition of small thicknesses of polymethyl methacrylate (PMMA) as a surrogate for the breast compression paddle, to produce a range of HVL values at each tube voltage. Using a spectral weighting method, D(g)N coefficients for the generated spectra were calculated for breast glandular densities of 0%, 12.5%, 25%, 37.5%, 50% and 100% for a range of compressed breast thicknesses from 3 to 8 cm. Eleven tables of normalized glandular dose (D(g)N) coefficients were produced for the following anode/filter combinations: W + 50 μm Ag, W + 500 μm Al, W + 700 μm Al, W + 200 μm Cu, W + 300 μm Cu, W + 50 μm Rh, Mo + 400 μm Cu, Mo + 30 μm Mo, Mo + 25 μm Rh, Rh + 400 μm Cu and Rh + 25 μm Rh. Where possible, these results were compared to previously published D(g)N values and were found to be on average less than 2% different than previously reported values.Over 200 pages of D(g)N coefficients were computed for modeled x-ray system spectra that are used in a number of new breast imaging applications. The reported values were found to be in excellent agreement when compared to published values.

Generation and analysis of clinically relevant breast imaging x‐ray spectra

Purpose The purpose of this work was to develop and make available x‐ray spectra for some of the most widely used digital mammography (DM), breast tomosynthesis (BT), and breast CT (bCT) systems in North America. Methods The Monte Carlo code MCNP6 was used to simulate minimally filtered (only beryllium) x‐ray spectra at 8 tube potentials from 20 to 49 kV for DM/BT, and 9 tube potentials from 35 to 70 kV for bCT. Vendor‐specific anode compositions, effective anode angles, focal spot sizes, source‐to‐detector distances, and beryllium filtration were simulated. For each 0.5 keV energy bin in all simulated spectra, the fluence was interpolated using cubic splines across the range of simulated tube potentials to produce spectra in 1 kV increments from 20 to 49 kV for DM/BT and from 35 to 70 kV for bCT. The HVL of simulated spectra with conventional filtration (at 35 kV for DM/BT and 49 kV for bCT) was used to assess spectral differences resulting from variations in: (a) focal spot size (0.1 and 0.3 mm IEC), (b) solid angle at the detector (i.e., small and large FOV size), and (c) geometrical specifications for vendors that employ the same anode composition. Results Averaged across all DM/BT vendors, variations in focal spot and FOV size resulted in HVL differences of 2.2% and 0.9%, respectively. Comparing anode compositions separately, the HVL differences for Mo (GE, Siemens) and W (Hologic, Philips, and Siemens) spectra were 0.3% and 0.6%, respectively. Both the commercial Koning and prototype “Doheny” (UC Davis) bCT systems utilize W anodes with a 0.3 mm focal spot. Averaged across both bCT systems, variations in FOV size resulted in a 2.2% difference in HVL. In addition, the Koning spectrum was slightly harder than Doheny with a 4.2% difference in HVL. Therefore to reduce redundancy, a generic DM/BT system and a generic bCT system were used to generate the new spectra reported herein. The spectral models for application to DM/BT were dubbed the Molybdenum, Rhodium, and Tungsten Anode Spectral Models using Interpolating Cubic Splines (MASMICSM‐T, RASMICSM‐T, and TASMICSM‐T; subscript “M‐T” indicating mammography and tomosynthesis). When compared against reference models (MASMIPM, RASMIPM, and TASMIPM; subscript “M” indicating mammography), the new spectral models were in close agreement with mean differences of 1.3%, −1.3%, and −3.3%, respectively, across tube potential comparisons of 20, 30, and 40 kV with conventional filtration. TASMICSbCT‐generated bCT spectra were also in close agreement with the reference TASMIP model with a mean difference of −0.8%, across tube potential comparisons of 35, 49, and 70 kV with 1.5 mm Al filtration. Conclusions The Mo, Rh, and W anode spectra for application in DM and BT (MASMICSM‐T, RASMICSM‐T, and TASMICSM‐T) and the W anode spectra for bCT (TASMICSbCT) as described in this study should be useful for individuals interested in modeling the performance of modern breast x‐ray imaging systems including dual‐energy mammography which extends to 49 kV. These new spectra are tabulated in spreadsheet form and are made available to any interested party.

NPS comparison of anatomical noise characteristics in mammography, tomosynthesis, and breast CT images using power law metrics

Digital mammography is the current standard for breast cancer screening, however breast tomosynthesis and breast CT (bCT) have been studied in clinical trials. At our institution, 30 women (BIRADS 4 and 5) underwent IRB-approved imaging by mammography, breast tomosynthesis, and bCT on the same day. Twenty three data sets were used for analysis. The 2D noise power spectrum (NPS) was computed and averaged for each data set. The NPS was computed for different slice thicknesses of dx × N, where dx ≈ 0.3 mm and N=1-64, on the bCT data. Each 2D NPS was radially averaged, and the 1D data were fit using a power law function as proposed by Burgess: NPS(f) = αf-β. The value of β was determined over a range of frequencies corresponding to anatomical noise, for each patient and each modality. Averaged over the 30 women (26 for bCT, 28 for tomosynthesis, 28 for mammography), for mammography β=3.06 (0.25), for CC tomosynthesis β=2.91 (0.35), and for axial bCT β=1.72 (0.47). For sagittal bCT β=1.77 (0.36) and for coronal bCT, β=1.88 (0.45). The computation of β versus slice thickness on the coronal bCT data set led to β≈1.7 for N=1, asymptotically reaching β ≈ 3 for larger slice thickness. These results suggest that there is a fundamental difference in breast anatomic noise as characterized by β, between thin slices (<2 mm) and thicker slices. Tomosynthesis was found to have anatomic noise properties closer to mammography than breast CT, most likely due to the relatively thick slice sensitivity profile of tomosynthesis.

Simulated lesion, human observer performance comparison between thin-section dedicated breast CT images versus computed thick-section simulated projection images of the breast.

The objective of this study was to compare the lesion detection performance of human observers between thin-section computed tomography images of the breast, with thick-section (>40 mm) simulated projection images of the breast. Three radiologists and six physicists each executed a two alterative force choice (2AFC) study involving simulated spherical lesions placed mathematically into breast images produced on a prototype dedicated breast CT scanner. The breast image data sets from 88 patients were used to create 352 pairs of image data. Spherical lesions with diameters of 1, 2, 3, 5, and 11 mm were simulated and adaptively positioned into 3D breast CT image data sets; the native thin section (0.33 mm) images were averaged to produce images with different slice thicknesses; average section thicknesses of 0.33, 0.71, 1.5 and 2.9 mm were representative of breast CT; the average 43 mm slice thickness served to simulate simulated projection images of the breast.The percent correct of the human observers responses were evaluated in the 2AFC experiments. Radiologists lesion detection performance was significantly (p < 0.05) better in the case of thin-section images, compared to thick section images similar to mammography, for all but the 1 mm lesion diameter lesions. For example, the average of three radiologists performance for 3 mm diameter lesions was 92% correct for thin section breast CT images while it was 67% for the simulated projection images. A gradual reduction in observer performance was observed as the section thickness increased beyond about 1 mm. While a performance difference based on breast density was seen in both breast CT and the projection image results, the average radiologist performance using breast CT images in dense breasts outperformed the performance using simulated projection images in fatty breasts for all lesion diameters except 11 mm. The average radiologist performance outperformed that of the average physicist observer, however trends in performance were similar. Human observers demonstrate significantly better mass-lesion detection performance on thin-section CT images of the breast, compared to thick-section simulated projection images of the breast.

SU‐C‐116‐02: Volumetric Dose Deposition in the Breast Using Real‐Time Dosimetry

PURPOSE To empirically determine the angular dose distribution to a breast in a digital breast tomosythesis and mammography system using real-time dosimetry technology. METHODS Three breast phantoms, constructed from polyethylene to simulate adipose tissue, were fabricated based on ninety cranio-caudal mammography patients. The smallest (3.2 cm), average (5.1 cm), and largest (7.0 cm) breast thickness matched the measured breast shape using 13 mm polyethylene sheets. Technique factors, which matched the automatic settings in both tomo and mammo were used. Phantoms were designed to be modular allowing a dose probe to be placed in various heights and depths. A total of 390 signal trains, comprised of 15-individual pulses (one for each angle) in tomo-mode and a single pulse for mammo, were collected using a real-time air ionization chamber (AccuGold, Radcal, Monrovia, CA). The signal train was integrated to generate a dose value at each location. This data was then used to generate a dynamic model of cumulative dose deposition as a function of x-ray tube angle in tomo and mammo modes. Additionally, the effective linear attenuation coefficient for each energy was calculated. Using a look up table for the linear attenuation of polyethylene, the effective energy was calculated. RESULTS The dose deposition decreases from the compression paddle to the detector. Cumulative dose from tomo scan was comparable to that of mammo; differences in cumulative dose increased with phantom size. The effective energy for tomo at 28, 31 and 35 kVp with Al filter was calculated to be 19.7, 21.7 and 25 kV respectively. For mammo at 26, 29 (Rh filter) and 30 kV (Ag filter) was found to be19.1, 20.4 and 23.7 respectively. CONCLUSION The volumetric dose deposition in Hologics dBT system shows angular dependence as a function of angle in tomo mode. NIH Grants.

Characterizing non-Gaussian properties of breast images with a noisy-Laplacian distribution

It is generally well known that the appearance of breast tissue in a mammogram is considerably more complex in a statistical sense than a simple random Gaussian texture, even when the correlation structure of the Gaussian has been set to match the power-law power spectrum of mammograms. However there has not been a systematic way to characterize the extent of departure from a Gaussian process. We address this topic here by proposing a noisy-Laplacian distribution to model response histograms derived from digital (or digitized) mammograms. We describe the distribution in terms of the probability density function and cumulative density function, as well as moments up to fourth order. We also demonstrate the usefulness of the new distribution by fitting it to responses from digital mammography.

SU‐D‐301‐05: Angular Dependency of the Slice Sensitivity Profile in Tomographic Breast Imaging

Purpose: To study the effect of angular acquisition on the slice sensitivity profile (SSP) in tomographic breast imaging, for both breast CT and tomosynthesis. Methods: Brass disks (0.025 mm thick) ranging in diameter from 2.5–20 mm were placed within mammography phantoms of varying thicknesses (3–6 cm). The disks were imaged using a prototype digital tomosynthesis system (Selenia Dimensions, Hologic Corporation, Bedford, Ma). The tomosynthesis system acquires images over an angular acquisition range of 15‐degrees. The same brass disks were used to measure the SSP in breast CT, using a 14cm‐diameter cylindrical polyethylene phantom. The bCT angular range spanned from 15 to 360‐degrees. A subset of the projection images were reconstructed with a filtered back projection algorithm to simulate limited‐angle acquisition. The SSP was determined by measuring the mean gray‐scale (GS) in the tomographic images along the z axis of the scan corresponding to the disk position. Normalized profiles of the background‐corrected mean GS values were generated as a function of the distance away from the detector. The full‐width at half maximum (FWHM) GS intensity was determined. Results: As the angular acquisition range increases, the SSP narrows and results in a smaller FWHM value. The FWHM values of the SSP for the smallest disk at (2.5 mm‐diameter) were 8.6 mm, 4.2 mm, and 0.8 mm for 15, 40, and 180 degrees, respectively. Conclusions: Limited angle tomography results in a broadening of the SSP, inversely proportional to the angular range of acquisition. The SSP approaches a minimum as the angular coverage exceeds 180 degrees. One author receives funding from Hologic (Bedford, Mass).