Takanori Tsunoo

Effect of scattered radiation on image noise in cone beam CT

Cone beam CT has a capability for the 3-dimensional imaging of large volumes with isotropic resolution, and has a potentiality for 4-dimensional imaging (dynamic volume imaging), because cone beam CT acquires data of a large volume with one rotation of an x-ray tube-detector pair. However, one of the potential drawbacks of cone beam CT is a larger amount of scattered x-rays, which may enhance the noise in reconstructed images, and thus affect the low-contrast detectablity. Our aim in this work was to estimate the scatter fractions and effects of scatter on image noise, and to seek methods of improving image quality in cone beam CT. First we derived a relationship between the noise in a reconstructed image and in an x-ray intensity measurement. Then we estimated the scatter to primary ratios in x-ray measurements using a Monte-Carlo simulation. From these we estimated the image noise under relevant clinical conditions. The results showed that the scattered radiation made a substantial contribution to the image noise. However, focused collimators could improve it by decreasing the scattered radiation drastically while keeping the primary radiation at nearly the same level. A conventional grid also improved the image noise, though the improvement was less than that of focused collimators.

Physical performance evaluation of a 256-slice CT-scanner for four-dimensional imaging.

We have developed a prototype 256-slice CT-scanner for four-dimensional (4D) imaging that employs continuous rotations of a cone-beam. Since a cone-beam scan along a circular orbit does not collect a complete set of data to make an exact reconstruction of a volume [three-dimensional (3D) image], it might cause disadvantages or artifacts. To examine effects of the cone-beam data collection on image quality, we have evaluated physical performance of the prototype 256-slice CT-scanner with 0.5 mm slices and compared it to that of a 16-slice CT-scanner with 0.75 mm slices. As a result, we found that image noise, uniformity, and high contrast detectability were independent of z coordinate. A Feldkamp artifact was observed in distortion measurements. Full width at half maximum (FWHM) of slice sensitivity profiles (SSP) increased with z coordinate though it seemed to be caused by other reasons than incompleteness of data. With regard to low contrast detectability, smaller objects were detected more clearly at the midplane (z = 0 mm) than at z = 40 mm, though circular-band like artifacts affected detection. The comparison between the 16-slice and the 256-slice scanners showed better performance for the 16-slice scanner regarding the SSP, low contrast detectability, and distortion. The inferiorities of the 256-slice scanner in other than distortion measurement (Feldkamp artifact) seemed to be partly caused by the prototype nature of the scanner and should be improved in the future scanner. The image noise, uniformity, and high contrast detectability were almost identical for both CTs. The 256-slice scanner was superior to the 16-slice scanner regarding the PSF, though it was caused by the smaller transverse beam width of the 256-slice scanner. In order to compare both scanners comprehensively in terms of exposure dose, noise, slice thickness, and transverse spatial resolution, K=Dsigma2ha3 was calculated, where D was exposure dose (CT dose index), sigma was magnitude of noise, h was slice thickness (FWHM of SSP), and a was transverse spatial resolution (FWHM of PSF). The results showed that the K value was 25% larger for the 16-slice scanner, and that the 256-slice scanner was 1.25 times more effective than the 16-slice scanner at the midplane. The superiority in K value for the 256-slice scanner might be partly caused by decrease of wasted exposure with a wide-angle cone-beam scan. In spite of the several problems of the 256-slice scanner, it took a volume data approximately 1.0 mm (transverse) x 1.3 mm (longitudinal) resolution for a wide field of view (approximately 100 mm long) along the zeta axis in a 1 s scan if resolution was defined by the FWHM of the PSF or the SSP, which should be very useful to take dynamic 3D (4D) images of moving organs.

Enlarged longitudinal dose profiles in cone-beam CT and the need for modified dosimetry

In order to examine phantom length necessary to assess radiation dose delivered to patients in cone-beam CT with an enlarged beamwidth, we measured dose profiles in cylindrical phantoms of sufficient length using a prototype 256-slice CT-scanner developed at our institute. Dose profiles parallel to the rotation axis were measured at the central and peripheral positions in PMMA (polymethylmethacrylate) phantoms of 160 or 320 mm diameter and 900 mm length. For practical application, we joined unit cylinders (150 mm long) together to provide phantoms of 900 mm length. Dose profiles were measured with a pin photodiode sensor having a sensitive region of approximately 2.8 x 2.8 mm2 and 2.7 mm thickness. Beamwidths of the scanner were varied from 20 to 138 mm. Dose profile integrals (DPI) were calculated using the measured dose profiles for various beamwidths and integration ranges. For the body phantom (320-mm-diam phantom), 76% of the DPI was represented for a 20 mm beamwidth and 60% was represented for a 138 mm beamwidth if dose profiles were integrated over a 100 mm range, while more than 90% of the DPI was represented for beamwidths between 20 and 138 mm if integration was carried out over a 300 mm range. The phantom length and integration range for dosimetry of cone-beam CT needed to be more than 300 mm to represent more than 90% of the DPI for the body phantom with the beamwidth of more than 20 mm. Although we reached this conclusion using the prototype 256-slice CT-scanner, it may be applied to other multislice CT-scanners as well.

Electron density measurement with dual-energy x-ray CT using synchrotron radiation.

Monochromatic x-ray computed tomography (CT) at two different energies provides information about electron density of human tissue without ambiguity due to the beam hardening effect. This information makes the treatment planning for proton and heavy-ion radiotherapy more precise. We have started a feasibility study on dual energy x-ray CT by using synchrotron radiation. A translation-rotation scanning CT system was developed for quantitative measurement in order to clarify what precision in the measurement was achieved. Liquid samples of solutions of K2HPO4 and solid samples of tissue equivalent materials were used to simulate human tissue. The experiments were carried out using monochromatic x-rays with energies of 40, 70 and 80 keV produced by monochromatizing synchrotron radiation. The solid samples were also measured in a complementary method using high-energy carbon beams to evaluate the electron densities. The measured electron densities were compared with the theoretical values or the values measured in the complementary method. It was found that these values were in agreement in 0.9% on average. Effective atomic numbers were obtained as well from dual-energy x-ray CT. The tomographic image based on each of the electron densities and the effective atomic number presents a different feature of the material, and its contrast drastically differs from that in a conventional CT image.

Magnitude and effects of x‐ray scatter in a 256‐slice CT scanner

We developed a prototype 256-slice CT scanner that employs continuous rotation of a cone-beam with a larger cone angle than conventional multidetector CTs (MDCT) to ensure a wide field of view. However, a larger cone angle may result in image deterioration due to increased x-ray scatter. Scattered radiation causes the detected signals to deviate from the true measurement of primary x-ray intensity and may result in artifacts (e.g., cupping and streak artifacts), quantitative inaccuracy in reconstructed CT number, and degradation of contrast-to-noise ratio (CNR). To reduce the effects of scatter, the 256-slice scanner incorporates an antiscatter collimator. Here, we estimated the magnitude of x-ray scatter in the prototype 256-slice CT scanner under clinical scan conditions and quantified the effects of this scatter on CT number accuracy, image noise, uniformity, and low contrast detectability. Although most experiments were performed with the antiscatter collimator, we also estimated the magnitude of x-ray scatter without the collimator to evaluate the scatter rejection efficiency of the collimator. The scatter-to-primary energy fluence ratio (SPR) without the collimator increased as cone angle increased, with estimated values of 49.7% for a 138 mm beam width with a phantom of 200 mm diameter, and 78.5% for a 320 mm diameter phantom. Estimated SPR was drastically decreased with the collimator, with an SPR reduction rate (ratio of SPR with and without the collimator) of 12.7% and 16.8% for the 200 and 320 mm diameter phantoms, respectively. The reduction in x-ray scatter by the collimator resulted in a considerable reduction in scatter effects. The measured uniformity was good and was independent of scatter amount. Although scatter still affected CT number accuracy, this could be corrected by rescaling. Further, although the CNR was decreased, in theory at least, the change was so subtle that it had no substantial effect on low-contrast detectability.

Properties of the prototype 256-row (cone beam) CT scanner.

We evaluated Feldkamp artifacts, which are specific to cone-beam computed tomography (CT), in phantom and clinical studies using the 256-multidetector-row CT (256MDCT), and compared the reconstruction accuracy of axial and helical scans. Image noise, slice sensitivity profile (SSP) and artifacts with the 256MDCT were evaluated using a phantom, and the results were compared to those of a 64MDCT. We also examined chest and abdomen scans produced with the 256MDCT in volunteers. For the axial scan, Feldkamp artifacts were visualized as high-frequency streak-like artifacts that are oriented horizontally at the edge of the scan region in the phantom study. Similar results were obtained with the volunteers in soft-tissue regions near either bony structures or air pockets. Feldkamp artifacts with the 256MDCT can lead to misdiagnosis if not correctly identified and minimized via helical scanning. Image noise was less for axial than helical scans, while SSP was better with helical than axial scans. Feldkamp artifacts observed in the 256MDCT images, however, did not generally affect the interpretation of images. The 256MDCT promises more accurate diagnosis, and will provide volumetric cine images of wider cranio-caudal coverage, enabling new applications of CT.

Design of synchrotron light source and its beamline dedicated to dual-energy x-ray computed tomography

A synchrotron light source dedicated to medical applications has been designed at National Institute of Radiological Sciences. The storage ring, with circumference of 80 m, is designed for acceleration of 2.3 GeV and a stored current of 420 mA. It is equipped with two multipole wigglers to produce sufficient photon flux in a hard x-ray region required for medical applications. The purposes of the synchrotron light source are clinical performance of medical diagnoses clinically and research and development relating with medical applications. One of the most interesting applications for us is dual-energy x-ray computed tomography (CT). It gives the information about electron density of human tissue. The information plays an important role in advancing heavy-ion radiotherapy of cancers. Electron density can be derived from attenuation coefficients measured by different energy x rays. In this paper, a practical method of the dual-energy x-ray CT with synchrotron radiation is proposed with the theoretical consideration. The primitive experiment using monochromatic x rays emitted from radioisotopes proved the procedure of analysis mentioned here effective to derive electron densities from linear attenuation coefficients for two x rays of a different energy. The beamline dedicated to dual-energy x-ray CT is also proposed. It has a multipole wiggler as a light source and it mainly consists of a dual crystal monochromator and a rotating filter for attenuating photon flux of x rays and two-dimensional detector.

Effect of scatter radiation on image noise in cone-beam CT

Cone beam CT has a capability of 3-dimensional imaging of large volumes with isotropic resolutions, and has a potentiality of 4-dimensional imaging (dynamic volume imaging) because cone beam CT acquires a large volume data with one rotation of X-ray tube-detector pair. However, one of the potential drawbacks with cone beam CT is larger amount of scattered X-rays. These X-rays may enhance the noise in reconstructed images, and thus affect low contrast detectability. The aim of this work was to estimate scatter fractions and effects of scatter on image noise, and was to seek methods of improving image quality in cone beam CT. First we derived a relationship between the noises in reconstructed image and in X-ray intensity measurement. Then we estimated scatter to primary ratios in X-ray measurements, using a Monte Carlo simulation. From these we estimated image noise in clinical relevant conditions. The results showed that scatter radiation made a substantial contribution to the image noise. However focused collimators improved it, because they decreased scatter radiation drastically while keeping the primary radiation nearly the same level. A conventional grid also improved image noise though the improvement was less than that of focused collimators.

Three dimensional segmented myocardial perfusion images by selective intracoronary injection of contrast using 256 slice cone beam computed tomography

Information of the segmented left ventricular (LV) area supplied by every coronary artery is potentially useful for strategy and practice of percutaneous coronary intervention, coronary artery bypass surgery, or percutaneous transluminal septal myocardial ablation (PTSMA).1,2 Myocardial contrast echocardiography (MCE) with intracoronary contrast injection has been used to evaluate the segmented LV area supplied by a coronary artery but it is invasive and evaluation of myocardial enhancement by ultrasound contrast material may depend upon the skill of the investigator.1 With a prototype high speed cone beam computed tomography (CT) apparatus employing 256 detector rows (Athena, Sony-Toshiba), an entire heart can be imaged within a single gantry rotation.3,4 Using this technique, combined with selective intracoronary injection of contrast, we determined the segmented LV area supplied by every coronary artery selectively. Two domestic pigs (20 kg each) were anaesthetised with isoflurane, and catheters positioned in the left anterior descending branch (LAD) of the coronary artery in pig 1 and the left circumflex branch (LCx) in pig 2, via the femoral arteries. The heart rate of pigs ranged between 70–80 beats per minute. Scan conditions for the 256 slice cone beam CT were: 120 kV, 200 mA, 1.0 second exposure, 1.0 second gantry rotation time, 0.5 mm slice thickness, and 256×0.5 mm slice collimation. For comparison, those for the 16 slice multislice CT scanner routinely used at our institute (SOMATOM Sensation 16, Siemens) were: …

Measurement of electron density in dual-energy x-ray CT with monochromatic x rays and evaluation of its accuracy.

Information on electron density is important for radiotherapy treatment planning in order to optimize the dose distribution in the target volume of a patient. At present, the electron density is derived from a computed tomography (CT) number measured in x-ray CT scanning; however, there are uncertainties due to the beam hardening effect and the method by which the electron density is converted from the CT number. In order to measure the electron density with an accuracy of +/-1%, the authors have developed dual-energy x ray CT using monochromatic x rays. They experimentally proved that the measured linear attenuation coefficients were only a few percent lower than the theoretical ones, which led to an accuracy within 2% for the electron density. There were three factors causing inaccuracy in the linear attenuation coefficient and the electron density: the influence of scattered radiation, the nonlinearity in the detector response function, and a theoretical process to derive the electron density from the linear attenuation coefficients. The linear attenuation coefficients of water were experimentally proved to differ by 1%-2% from the theoretical one even when the scattering effect was negligible. The nonlinearity of the response function played an important role in correcting the difference in the linear attenuation coefficient. Furthermore, the theoretical process used for deriving the electron density from the linear attenuation coefficients introduces about 0.6% deviation from the theoretical value into the resultant electron density. This deviation occurs systematically so that it can be corrected. The authors measured the electron densities for seven samples equivalent to soft tissue in dual-energy x-ray CT, and finally obtained them with an accuracy of around +/-1%.