Typical values of z-resolution for different Digital Breast Tomosynthesis systems evaluated in a multicenter study

PURPOSE

The aim of the present study, conducted by a working group of the Italian Association of Medical Physics (AIFM), was to define typical z-resolution values for different digital breast tomosynthesis (DBT) models to be used as a reference for quality control (QC). Currently, there are no typical values published in internationally agreed QC protocols.

METHODS

To characterize the z-resolution of the DBT models, the full width at half maximum (FWHM) of the artifact spread function (ASF), a technical parameter that quantifies the signal intensity of a detail along reconstructed planes, was analyzed. Five different commercial phantoms, CIRS Model 011, CIRS Model 015, Modular DBT phantom, Pixmam 3-D, and Tomophan, were evaluated on reconstructed DBT images and 82 DBT systems (6 vendors, 9 models) in use at 39 centers in Italy were involved.

RESULTS

The ASF was found to be dependent on the detail size, the DBT angular acquisition range, the reconstruction algorithm and applied image processing. In particular, a progressively greater signal spread was observed as the detail size increased and the acquisition angle decreased. However, a clear correlation between signal spread and angular range width was not observed due to the different signal reconstruction and image processing strategies implemented in the algorithms developed by the vendors studied.

CONCLUSIONS

The analysis led to the identification of typical z-resolution values for different DBT model-phantom configurations that could be used as a reference during a QC program.

Objective measurements of image quality in synchrotron radiation phase-contrast imaging versus digital mammography

PURPOSE

Phase-contrast mammography with synchrotron radiation is an innovative X-ray imaging practice that improves the identification of breast lesions. Previous studies have proven the superiority of the mammography images taken in the phase-contrast modality using synchrotron radiation beams as compared with images taken in conventional mammography by subjective analyses. However, to our knowledge, no previous study has compared different acquisition systems in order to quantify this improvement by means of objective robust indicators. In this research, we intend to quantify the superiority of phase-contrast imaging by means of objective metrics of image quality.

METHODS

Images from the American College of Radiology Mammographic Accreditation Phantom were obtained at hospitals, in two digital mammography equipment and at the Elettra synchrotron radiation facility (Trieste, Italy), using free space propagation phase-contrast modality. Regions of interest were selected to analyze image quality at the fibers (phase object) and masses (area object) simulated on the phantom by means of the signal-to-noise ratio, the figure of merit, the contrast and the edge visibility.

RESULTS

The image contrast and edge visibility were significantly higher at the phase-contrast modality as compared with digital mammography equipment. The figure of merit using phase-contrast modality was higher for the fibers and comparable for the masses.

CONCLUSION

The results showed an improvement of the contrast and edge visibility in phase-contrast images. These improvements may be important in the detection of small lesions and details.

The mammography project at the SYRMEP beamline

A clinical program for X-ray phase contrast (PhC) mammography with synchrotron radiation (SR) has been started in March 2006 at the SYRMEP beamline of Elettra, the SR facility in Trieste, Italy. The original beamline layout has been modified substantially and a clinical facility has been realized. In order to fulfill all security requirements, dedicated systems have been designed and implemented, following redundancy criteria and "fail safe" philosophy. Planar radiographic images are obtained by scanning simultaneously the patient and the detector through the stationary and laminar SR beam. In this first phase of the project a commercial screen-film system has been used as image receptor. Upon approval by the respective authorities, the mammography program is about half way to conclusion. Up to now about 50 patients have been examined. The patients are volunteers recruited by the radiologist after conventional examinations at the hospital resulted in an uncertain diagnosis. As an example one case of PhC SR mammography is shown and compared to conventional digital mammography. Preliminary analysis shows the high diagnostic quality of the PhC SR images that were acquired with equal or less delivered dose compared to the conventional ones.

Evaluation of image quality and dose in renal colic: comparison of different spiral-CT protocols

The aim of this study was to test different technical spiral-CT parameters to obtain optimal image quality with reduced X-ray dose. Images were acquired with a spiral-CT system Philips Tomoscan AVE1, using 250 mA, 120 kV, and 1-s rotational time. Three protocols were tested: protocol A with 5-mm thickness, pitch 1.6, slice reconstruction every 2.5 mm; protocol B with 3-mm thickness, pitch 1.6, slice reconstruction every 1.5 mm; and protocol C with 3-mm thickness, pitch 2, slice reconstruction every 1.5 mm. Two phantoms were employed to evaluate the image quality. Axial images were acquired, then sagittal and coronal images were reconstructed. Finally, the absorbed X-ray dose for each protocol was measured. Regarding image quality, 5-mm-thick images (protocol A) showed greater spatial resolution and lower noise compared with 3-mm-thick images (protocols B and C) on the axial plane; 3-mm reconstructed sagittal and coronal images (protocols B and C) showed an improved image quality compared with 5-mm reformatted images (protocol A). Concerning X-ray dose, the mean dose was: protocol A 19.6 +/- 0.8 mGy; protocol B 14.4 +/- 0.6 mGy; protocol C 12.5 +/- 1.0 mGy. Our study supports the use of thin slices (3 mm) combined with pitch of 1.6 or 2 in renal colic for X-ray dose reduction to the patient and good image quality.

[Effective dose in X-ray examinations: comparison between spiral CT and urography in the study of renal colic]

AIM

The aim of this study was to calculate the effective dose in patients with renal colic undergoing spiral CT examination and IVU examination, and to verify the ability of the systems to measure the effective dose.

MATERIAL AND METHODS

Dose measurements were carried out for both diagnostic techniques by employing an anthropomorphic phantom with thermoluminescence detectors (TLDs) fitted inside the phantom. The data so obtained were compared with dose estimations based on published tables, which allow to obtain the effective dose on the basis of the experimental value of common dose indicators, the performance of the equipment and the setting of the acquisition parameters for such examinations.

RESULTS

The absorbed dose inside the phantom had an homogeneous distribution during the spiral CT examination, while during the IVU there were significant differences in the absorbed dose between different zones of the body, due to the geometry of the x-ray beam. The mean effective dose, which corresponds to the average of values absorbed by males and females, measured by TLD dosimeters for spiral CT examination was 3.3 time that for IVU. For both spiral CT and IVU the dose to the male was significantly lower than that to the female, owing to the anatomic position of female gonads that are completely exposed to the x-ray radiations during the study, while male gonads lie outside of the exposed area. Both methods for estimating the effective dose in spiral CT, based on CTDI value, significantly underestimated the value derived from experimental TLDs measurement. The evaluation of effective dose in IVU, based on the measurement of the skin dose in air, overestimated the measurement performed by TLDs.

DISCUSSION AND CONCLUSIONS

Dose measurement performed by TLD dosimeters fitted inside an anthropomorphic phantom has the advantage of calculating experimentally the absorbed dose in different anatomic districts. However, the method is quite difficult, and introduces some imprecisions due to the simplified morphology of the phantom and the measuring system of TLDs (+/- 10%). Another possibility is to make theoretical estimations on the basis of simple dose measurements. Also these methods are affected by many factors that may introduce imprecisions, such as the patient geometry which has a considerable effect on the dose distribution inside the body. Therefore the methods of measurement based on CTDI value can provide only rough estimations of effective dose. The same considerations can be applied to IVU; in this case the theoretical estimations are based on skin dose measurements, which are very accurate but cannot foresee the behaviour of x-rays inside the patient. As regards the diagnostic procedures evaluated, the gonadic absorbed dose has a greater effect on the effective dose value as a consequence of the weight tissue ratio Wt. In this respect we should underline that, while female gonads are completely irradiated during spiral CT and IVU examinations, the male gonads are exposed only to diffuse radiation; this entails a great difference between the effective dose to the male and that to the female. For this reason the effective dose should be expressed not only for the average man (that is the mean value between males and females), but also distinguishing by patient sex. Although the comparison of different evaluating methods confirmed the difficulties in obtaining the exact effective dose, all the measurements performed with IVU and spiral CT showed that the effective dose for CT is 3.3 times that for IVU. However, the excellent image quality obtained by spiral CT and the high intrinsic contrast of stones probably makes it possible to reduce the dose with this technique, preserving an acceptable quality of the diagnostic images.

Wiener spectrum of radiographic systems: comparison of different evaluation methods

The noise power spectrum, or Wiener spectrum, of the radiographic mottle is a fundamental quantity in film-screen image quality evaluation. In this paper, using a high-quality computerized microdensitometer, two different acquisition and calculation methods for noise evaluation are compared. The first one is the classic (unidimensional) method used in film noise evaluation: a long and narrow slit (10 x 400 microns2) is used to delimit the microdensitometer light beam and the transmission data are collected by scanning the sample in a rectilinear pattern. A section of the two-dimensional Wiener spectrum is thus obtained. The second (two-dimensional) method is similar to that used in digital image noise evaluation: a square slit is used on the microdensitometer window and data are collected by scanning the sample on a square pattern. To evaluate the effect of different sampling frequencies, our data were acquired both selecting a 50 x 50 microns2 square slit and a 20 x 20 microns2 square slit. The two-dimensional Wiener spectrum thus obtained is then reduced to a unidimensional function. The measurements were made on two different films (Kodak Ortho G e Kodak T-MAT G) exposed with the same screen (Kodak Lanex Regular). These films have the same sensitivity but a different emulsion structure. One film (Ortho G) is made of irregular halide silver grains and the other (T-MAT G) of tabular grains. A satisfactory agreement between the two procedures was found which makes the comparison of data from the laboratories using microdensitometers and those using TV-grabbing system for film-screen evaluation meaningful.

[The dose and quality program in mammography (DQM). Results of the study carried out in 20 Fruili-Venezia Giulia centers]

Recently, 20 breast centers in Friuli Venezia Giulia have been investigated for the mammography optimization program known as DQM (Dose and Quality in Mammography). This was the second trial, the first one dating 1987-1988 and including only 10 centers. The chosen parameters were evaluated by means of a performance phantom containing 15 details, a Victoreen 660 ionizing camera, a star pattern and several plexiglass phantoms of different thickness. Statistical-methodological data were collected by means of a questionnaire. The results of the trial were communicated to each center as well as directions for future improvements. All centers were found to use mammographic equipment with a molybdenum anode and the film-screen combination; 18 center use grids daily, while 11 can perform direct radiographic magnification. Only 40% of the examined centers use a dedicated daylight printer. The overall quality of the radiographic image was seen to be acceptable, as 8 or more details of 15 hidden in the phantom were visualized, in 16 of 20 centers. The mean glandular dose was below the minimum threshold in all centers: in 45% of cases it was below 1.5 mGy. The setting of the automatic exposition control system was found to be poor, especially that of the adjustment related to breast thickness. Acceptability thresholds were met only by 31% of the examined centers (density variations < 20%). Notwithstanding the identification of areas where improvements can be made also from a methodological point of view, we can conclude that the quality provided by regional breast centers is definitely adequate. Good image quality and acceptable doses were observed in 80% of the centers. The results are significant and encouraging when compared with those from the first DQM trial carried out in our region, which confirms the value of periodical quality programs.

The choice of radiographic film-screen systems: quality evaluation and purchasing specifications drafting. Report on A.I.F.B. Working Group activity

The comparative evaluation of radiographic screen-film systems presents several problems from both the theoretical and the experimental points of view. From the theoretical point of view the main difficulties are related to the choice of the parameters best suited to express the "overall quality" of a system. This quantity is expressed as a product of image quality index and system sensitivity. As image quality index we assumed the signal-to-noise power ratio: this index depends in an explicit way on contrast, resolution and noise of the system. From the experimental point of view the main problem is that to measure some basic quantities, sophisticated and expensive equipment, like computer-controlled microdensitometers, is generally required. In this paper, we report the Italian Association of Biomedical Physicists Task group suggestions for measuring the basic physical parameters (with particular reference to the use of cost-effective equipment and for purchasing specification drafting). Using synthetic quality indices, the evaluation criteria of radiographic materials are directly derived from the general theory of radiographic image perception.