bCT

The aim of this project is to study the irradiation conditions, the absorbed dose and image quality in X-ray Breast Computed Tomography (BCT), using a laboratory CT scanner assembled ad hoc and basic tomographic techniques, whose realization was started by the proposers in the framework of a previous project supported by INFN. The following scientific tasks of investigation have been identified: 1. - Analysis of dose distribution In this project we want to measure the absorbed radiation dose delivered to breast phantoms by our bCT laboratory scanner, using a PMMA breast phantom in which TLDs were positioned both along the axis and at the phantom periphery, i.e. toward the "breast skin". Dose will be measured with a hemiellipsoid PMMA phantom of 14-cm maximum diameter, for a fixed milliampere-second (mAs) at tube voltages ranging from 40 to 80 kVp. Our investigative goals are: i) To estimate (in lucite) the dose drop radially from the surface to the central layer, and axially from the "nipple" to the "chest wall". These results could represent the experimental support to simulation analysis of dose distribution in uncompressed breast model. In these simulations, the dose distribution values due to different materials (PMMA and breast tissue) can be studied and the experimental data, obtained with PMMA, can be compared. ii) Analysis of dose distribution due to different beam energy characteristics obtained by varying the X-ray tube added filtration (from 0 to 300 m Cu). iii) Analysis and comparison of measurements performed by the use of synchrotron radiation at the ELETTRA Laboratory, where there is the dedicated beam line for mammography (SYRMEP) and for breast tomography. Data give indication of a lower dose to the organ in synchrotron radiation monochromatic parallel beam with respect to CBBCT. At ELETTRA, we plane to obtain beam time in order to measure the dose in PMMA phantoms with monochromatic radiation in the range 20-28 keV. A proposal to ELETTRA Scientific Commettee has been accepted for a shift at SYRMEP during July 2008 and November 2009 for tests. iv) The breast thickness influences the dose delivered to the organ and the MDG. So it is very important to establish the standard dimensions of an uncompressed breast. To this aim, Boone et al., considering a sample of 200 north-american women, showed that the mean diameter of the uncompressed breast is 14.0 cm. Then they correlated the uncompressed breast diameter to the compressed breast thickness, according to which a diameter of 14 cm corresponds to a compressed thickness of 5 cm during a conventional X-ray mammography exam. We plan to measure in a large group of women the distribution of breast diameter at chest wall and corresponding compressed breast thickness, in order to reduce the average biometric data for a population of women. A first sample of 100 women has been analyzed at the National Cancer Institute in Napoli, showing an average breast diameter of 12.2. cm. This analysis is relevant to dosimetric evaluation in bCT. 2.- Tumor Masses visibility In a recent paper the first report has been published about the clinical test of a CBBCT apparatus in a number of female patients, for which the standard screen-film X-ray mammography procedure (two-views) was performed. This group involved both healthy and cancer patients. An overall score of mammography vs. CBBCT for each patient was obtained, in terms of diagnostic quality in cancer diagnosis, relative to lesion visualization. As regards the visibility of tumor masses, breast CT was reported to perform significantly better than mammography for visualization of masses. It can be speculated that the better visibility of masses in bCT arises from the inherent tomographic nature of CT imaging. In fact, the blurring effect of overlying tissue structures on the visibility of masses that occurs in the planar views of conventional mammography, is not present in the slice image of the tissue lesions, whose structure is contrasted against the neighboring tissue in each tissue slice. Yet, it is not well known what is the minimum tissue contrast of masses visible in the bCT exam. Moreover, possible ways could be explored in order to increase the inherent low contrast of cancer tissue masses with respect to glandular/adipose tissue, without resorting to contrast agent imaging. With our laboratory bCT scanner, we plan to investigate in PMMA phantoms the visibility of masses, simulated as small (from 5 up to 15 mm diameter) spheres embedded in PMMA. The spheres will be made of plastics (PMMA, PC, Delrin, Teflon, PE). Fat tissue-like details will be simulated, too. Particular interest in this respect, is the possibility envisaged of increasing the mass contrast using phase-contrast CT imaging. In this X-ray imaging modality, which requires the use of synchrotron radiation or of microfocus (<20 micron) focal spot sizes of the X-ray tube, an edge-enhancment effect is produced by the diffraction of X-rays at the border of material discontinuities (as opposed to absorption of X-rays in the material bulk) and interference of the diffracted rays after transmission through the detail. It is our investigative guess that phase-contrast effect in bCT could be advantageously used to improve even further the visibility of breast tissue details even in so-called dense breasts, where detail visibility is impaired. To this purpose, the mini-focus X-ray source available on the CT scanner will be replaced by a microfocus X-ray source as contrast visibility studies will be performed using PMMA phantoms. 3.- Microcalcifications visibility Microcalcifications are small calcium deposits within the breast that might be early indicators of breast cancer. The size of microcalcifications varies, ranging from a fraction of mm to less than 100 μm. They are organized as a single spot or as ‘cluster'. The ‘cluster' of microcalcifications is defined by the identification of five or more microcalcifications in a volume of a cubic centimetre. The size, the shape, the structure and the number of microcalcifications present in a cluster are factors that are the radiologist evaluates in oreder to recognize and classifies a lesion; in particular the radiologist is able to associate all morphological characteristics of microcalcifications to a grade of the malignant lesion. It is expected that breast CT should improve the microcalcification detectability due to the fact that the 3D visualization removes the limitation due to the superposition of breast structures present in the 2D mammography. Simulation studies have been performed in order to evaluate what size microcalcifications would be detectable with flat panel breast CT and how this size changes with the detector pixel dimensions and with the mean glandular dose. The results indicate that the mormalized area under the ROC (Receiver Operating Characteristics) for detecting the microcalcification lesions with diameter larger than 0.175 mm is grater that 0.95 for detector pixel size of 0.1 mm. A previously experimental study reports that 0.2 mm calcifications (spheres embedded in a phantom) were visible with a 0.18 mm pixel size detector. More recent studies conducted on breast phantoms in order to investigate the effect of radiation dose indicated that the microcalcification visibility increase with the mean glandular dose and decrease with the breast size. First studies on patients indicate that the screen-film mammography currently appears to be better than CT for imaging microcalcifications lesions. This scenario - according to which there is no exhaustive response about the microcalcification visibility/detectability in breast CT - suggests that an improvement of the spatial resolution in the breast CT images should be required. So we propose to study the microcalcification visibility/detectability with our CBBCT prototype. Our set-up presents two peculiar features with respect to the other breast CT systems previously described. First, the detector pixel size is lower (50 micron instead of about 400 micron) and second, the X-ray tube focal spot size is lower (50 micron instead of about 400 micron), as compared with the U.C. Davis apparatus. In fact, , currently, in breast CT the utilized detectors have a pixel size > 200 micron and the focal spot size of the X-ray tube is in the order of several micrometers (400-600 micron). In particular, using phantoms (see previous sections) with insert simulating microcalcification lesions (single and cluster), we intend to evaluate, in terms of contrast and spatial resolution, the microcalcifications visibility/detectability as a function of the detector pixel size (50-100-200 micron); the X-ray technique (kVp, filtration, focus size, mAs). 4.- SRBCT and CBBCT techniques comparison Two irradiation modalities (synchrotron radiation monochromatic parallel beam, SRBCT, and polychromatic X-ray cone beam, CBBCT) are used in BCT as recently proposed the SYRMEP group led by Prof. E. Castelli in Trieste and by the Napoli research group. SRBCT as performed at ELETTRA facility in Trieste uses monochromatic X-ray beams in the energy range 20-28 keV with single photon counting silicon detectors. The SRBCT and CBBCT techniques are in fact complementary: one is based on a parallel beam while the other on a divergent beam, one is monochromatic while the other polychromatic. So these techniques allow to investigate many irradiation conditions. At present there is no study - according to the proposers's best knowledge – which provides a comparison of these two techniques. Monochromatic irradiation produces no beam hardening artefacts in the CT reconstruction, and use of low-energy X-rays (e.g., 28 keV) is expected to determine a minor influence of X-ray Compton scattering in the tissue volume on the image quality, as compared to polychromatic X-ray beam at effective energies in the 40-55 keV range produced by 80 kVp tube voltages. In addition, it is not clear whether a low-energy irradiation in BCT allows for a reduction in Mean Glandular Dose with respect to high-energy irradiation. A research effort will be dedicated to assess the effective performance of the two BCT systems with a technological and methodological comparison on the same breast phantoms. In particular, it is relevant to this project the study of the adsorbed dose, the dose distribution and the radiation scattering influence on the image quality. At the end of the project, the expected results will be the assessing of two BCT techniques intrinsic potentialities with regard to the different irradiation geometries, the beam monochromaticity, the beam energy, the beam geometries and the scattering reduction.