Investigation of the spatial resolution of an online dose verification device.

PURPOSE: The aim of this work is to characterize a new online dose verification device, COMPASS transmission detector array (IBA Dosimetry, Schwarzenbruck, Germany). The array is composed of 1600 cylindrical ionization chambers of 3.8 mm diameter, separated by 6.5 mm center-to-center spacing, in a 40 × 40 arrangement. METHODS: The line spread function (LSF) of a single ion chamber in the detector was measured with a narrow slit collimator for a 6 MV photon beam. The 0.25 × 10 mm(2) slit was formed by two machined lead blocks. The LSF was obtained by laterally translating the detector in 0.25 mm steps underneath the slit over a range of 24 mm and taking a measurement at each step. This measurement was validated with Monte Carlo simulation using BEAMnrc and DOSXYZnrc. The presampling modulation transfer function (MTF), the Fourier transform of the line spread function, was determined and compared to calculated (Monte Carlo and analytical) MTFs. Two head-and-neck intensity modulated radiation therapy (IMRT) fields were measured using the device and were used to validate the LSF measurement. These fields were simulated with the BEAMnrc Monte Carlo model, and the Monte Carlo generated incident fluence was convolved with the 2D detector response function (derived from the measured LSF) to obtain calculated dose. The measured and calculated dose distributions were then quantitatively compared using χ-comparison criteria of 3% dose difference and 3 mm distance-to-agreement for in-field points (defined as those above the 10% maximum dose threshold). RESULTS: The full width at half-maximum (FWHM) of the measured detector response for a single chamber is 4.3 mm, which is comparable to the chamber diameter of 3.8 mm. The pre-sampling MTF was calculated, and the resolution of one chamber was estimated as 0.25 lp∕mm from the first zero crossing. For both examined IMRT fields, the χ-comparison between measured and calculated data show good agreement with 95.1% and 96.3% of in-field points below χ of 1.0 for fields 1 and 2, respectively (with an average χ of 0.29 for IMRT field 1 and 0.24 for IMRT field 2). CONCLUSIONS: The LSF for a new novel online detector has been measured at 6 MV using a narrow slit technique, and this measurement has been validated by Monte Carlo simulation. The detector response function derived from line spread function has been applied to recover measured IMRT fields. The results have shown that the device measures IMRT fields accurately within acceptable tolerance.

Physical characterization and performance comparison of active- and passive-pixel CMOS detectors for mammography.

We investigated the physical characteristics of two complementary metal oxide semiconductor (CMOS) mammography detectors. The detectors featured 14-bit image acquisition, 50 microm detector element (del) size and an active area of 5 cm x 5 cm. One detector was a passive-pixel sensor (PPS) with signal amplification performed by an array of amplifiers connected to dels via data lines. The other detector was an active-pixel sensor (APS) with signal amplification performed at each del. Passive-pixel designs have higher read noise due to data line capacitance, and the APS represents an attempt to improve the noise performance of this technology. We evaluated the detectors' resolution by measuring the modulation transfer function (MTF) using a tilted edge. We measured the noise power spectra (NPS) and detective quantum efficiencies (DQE) using mammographic beam conditions specified by the IEC 62220-1-2 standard. Our measurements showed the APS to have much higher gain, slightly higher MTF, and higher NPS. The MTF of both sensors approached 10% near the Nyquist limit. DQE values near dc frequency were in the range of 55-67%, with the APS sensor DQE lower than the PPS DQE for all frequencies. Our results show that lower read noise specifications in this case do not translate into gains in the imaging performance of the sensor. We postulate that the lower fill factor of the APS is a possible cause for this result.

A bench-top megavoltage fan-beam CT using CdWO4-photodiode detectors. I. System description and detector characterization.

We describe the components of a bench-top megavoltage computed tomography (MVCT) scanner that uses an 80-element detector array consisting of CdWO4 scintillators coupled to photodiodes. Each CdWO4 crystal is 2.75 x 8 x 10 mm3. The detailed design of the detector array, timing control, and multiplexer are presented. The detectors show a linear response to dose (dose rate was varied by changing the source to detector distance) with a correlation coefficient (R2) nearly unity with the standard deviation of signal at each dose being less than 0.25%. The attenuation of a 6 MV beam by solid water measured by this detector array indicates a small, yet significant spectral hardening that needs to be corrected before image reconstruction. The presampled modulation transfer function is strongly affected by the detector's large pitch and a large improvement can be obtained by reducing the detector pitch. The measured detective quantum efficiency at zero spatial frequency is 18.8% for 6 MV photons which will reduce the dose to the patient in MVCT applications. The detector shows a less than a 2% reduction in response for a dose of 24.5 Gy accumulated in 2 h; however, the lost response is recovered on the following day. A complete recovery can be assumed within the experimental uncertainty (standard deviation <0.5%); however, any smaller permanent damage could not be assessed.

Size and positioning reproducibility of an 192Ir brachytherapy stepping source.

In this paper we describe techniques for measuring the dimensions and position reproducibility of an 192Ir brachytherapy stepping source. Measurements were carried out using a 0.25x10x152 mm3 collimator placed in front of a detector of our own design. The brachytherapy source was translated past the collimator in 0.025 mm increments using a stepper motor. The source was found to be 3.58 mm long and 0.69 mm wide, which is in good agreement with the manufacturer's values of 3.5x0.6 mm2. The source position was reproducible to within 0.12 mm.

The axolotl as an animal model for the comparison of 3-D ultrasound with plain film radiography.

We assessed the usefulness of an animal model, the axolotl (Ambystoma mexicanum), in comparing 3-D ultrasound (3-D US) and plain film radiographs. Hindlimbs were amputated from 5 animals, at either the zeugopodial or stylopodial level, and each regenerating limb was imaged 16 times with 3-D US and 14 times with plain film X ray over 315 days. US images were acquired with a Siemens Sonoline Versa Pro and a 10-MHz linear array transducer. For 3-D US images, the probe was translated in a motor-driven linear stage while images were digitized. The regenerating tibia and fibula bones were detected on 3-D US an average of 37 days earlier than on plain film radiography, and regenerating phalangeal bones were detected on 3-D US an average of 18 days earlier. After 120 days, both imaging modalities consistently showed the bones. The average bone growth rates for the tibia and fibula were 0.019 +/- 0.001 mm/day and 0.019 +/- 0.001 mm/day, respectively.

A Doppler ultrasound clutter phantom.

We describe two variations of a phantom designed to evaluate the wall filters implemented on colour and spectral Doppler instruments. Both variations use an acoustic beam splitter to place the same Doppler sample volume within a motor-driven clutter belt and a flow source, which is either a second belt (dual-belt phantom) or a vascular phantom (belt/vascular phantom). We used the dual-belt phantom to evaluate the effects of the clutter belt velocity, flow belt velocity and clutter-to-flow power ratio on the reported colour Doppler shifts. The results show that the choice of wall filter, as well as the amplitudes and velocities of the clutter and flow components, affect the measured Doppler shifts. Results obtained with the belt/vascular phantom show that colour Doppler shifts due to the moving fluid depend strongly on the clutter velocity and choice of wall filter. However, only a small dependence on Doppler signal strength was observed.

Evaluation of an automated real-time spectral analysis technique.

An adaptive real-time Doppler peak-frequency tracing algorithm was evaluated in vitro and compared to manual peak-frequency traces. A computer-controlled pump was used to generate physiological flow waveforms in a vasculature-mimicking phantom. Spectral waveforms were obtained on an ATL HDI along with real-time estimates of diagnostic parameters, including maximum systolic, minimum diastolic, time-averaged peak frequencies and pulsatility and resistance indices. The effect of the signal-to-noise ratio on the measured parameters was investigated. The imprecision in the measured parameters was found to depend somewhat on the waveform shape; e.g., the imprecision in PI was 4.1% for a normal renal waveform and 8.5% for a waveform having reverse diastolic flow. The peak frequency envelopes of the same waveform data were traced manually by nine operators, and the resulting diagnostic parameters were compared to ones obtained from automated peak-frequency traces of the same waveform data. The agreement between parameters measured by the automated routine and those measured manually was found to depend somewhat on the waveform shape; e.g., the bias in the PI was 1.3% for a renal waveform lacking diastolic flow, and 12% for a waveform with reverse diastolic flow. The between-observer variations in the manual measurements ranged from 0.8% up to 9.4%. The overall variations associated with the automated traces were found to be smaller than or equal to those of the manual traces.

Quantitative investigation of in vitro flow using three-dimensional colour Doppler ultrasound.

A quantitative in vitro flow study was performed by using a three-dimensional colour Doppler imaging system. This system was based on a clinical ultrasound instrument with its transducer mounted on a motor-driven translation stage. A vascular and tissue-mimicking phantom containing two wall-less vessels, one normal and another stenotic, was used to quantify the measurement accuracy of the flow velocity and the flow field. Steady state flows, having Reynolds numbers ranging between 460 and 1300, were generated by a computer-controlled positive displacement pump. Effects of the parameter settings of the ultrasound instrument on results of the estimation of flow field were also studied. Experimental results show that our three-dimensional colour Doppler system's velocity accuracy was better than 7% of the Nyquist velocity and its spatial accuracy was better than 0.5 mm. The system showed a good correlation (r = 0.999) between the estimated and the true mean flow velocity, and a good correlation (r = 0.998) between the estimated maximum and the true mean flow velocity. This study is our first step toward validating the measurement of the three-dimensional velocity and wall shear stress distributions by using three-dimensional colour Doppler ultrasound

A wall-less vessel phantom for Doppler ultrasound studies.

Doppler ultrasound flow measurement techniques are often validated using phantoms that simulate the vasculature, surrounding tissue and blood. Many researchers use rubber tubing to mimic blood vessels because of the realistic acoustic impedance, robust physical properties and wide range of available sizes. However, rubber tubing has a very high acoustic attenuation, which may introduce artefacts into the Doppler measurements. We describe the construction of a wall-less vessel phantom that eliminates the highly attenuating wall and reduces impedance mismatches between the vessel lumen and tissue mimic. An agar-based tissue mimic and a blood mimic are described and their acoustic attenuation coefficients and velocities are characterised. The high attenuation of the latex rubber tubing resulted in pronounced shadowing in B-mode images; however, an image of a wall-less vessel phantom did not show any shadowing. We show that the effects of the highly attenuating latex rubber vessels on Doppler amplitude spectra depend on the vessel diameter and ultrasound beam width. In this study, only small differences were observed in spectra obtained from 0.6 cm inside diameter thin-wall latex, thick-wall latex and wall-less vessel phantoms. However, a computer model predicted that the spectrum obtained from a 0.3-cm inside diameter latex-wall vessel would be significantly different than the spectrum obtained from a wall-less vessel phantom, thus resulting in an overestimation of the average fluid velocity. These results suggest that care must be taken to ensure that the Doppler measurements are not distorted by the highly attenuating wall material. In addition, the results show that a wall-less vessel phantom is preferable when measuring flow in small vessels.

A geometrically accurate vascular phantom for comparative studies of x-ray, ultrasound, and magnetic resonance vascular imaging: construction and geometrical verification.

A technique for producing accurate models of vascular segments for use in experiments that assess vessel geometry and flow has been developed and evaluated. The models are compatible with x-ray, ultrasound, and magnetic resonance (MR) imaging systems. In this paper, a model of the human carotid artery bifurcation, is evaluated that has been built using this technique. The phantom consists of a thin-walled polyester-resin replica of the bifurcation through which a blood-mimicking fluid may be circulated. The phantom is surrounded by an agar tissue-mimicking material and a series of fiducial markers. The blood- and tissue-mimicking materials have x-ray, ultrasound, and MR properties similar to blood and tissue; fiducial markers provide a means of aligning images acquired by different modalities. The root-mean-square difference between the inner wall geometry of the constructed model and the desired dimensions was 0.33 mm. Static images were successfully acquired using x-ray, ultrasound, and MR imaging systems, and are free of significant artifacts. Flow images acquired with ultrasound and MR agree qualitatively with each other, and with previously published flow patterns. Volume-flow measurements obtained with ultrasound and MR were within 4.4% of the actual values.

Three-dimensional colour Doppler imaging.

We have developed a system to acquire in vivo three-dimensional (3D) colour velocity images of peripheral vasculature. A clinical ultrasound system was modified by mounting the transducer on a motor-driven translation stage, allowing planar ultrasound images to be acquired along a 37 mm long stroke. A 3D velocity image is acquired by digitizing, in synchrony with the cardiac cycle, successive video images as the transducer is moved over the skin surface. 3D images require about 1 min to acquire and 10 min to reconstruct before being viewed interactively. Image acquisition at several points in the cardiac cycle permits a cine-type reconstructed image. Geometrical, temporal and velocity accuracy of the acquisition and reconstruction have been quantified and found not to degrade the image.

Computer-controlled flow simulator for MR flow studies.

A novel computer-controlled flow simulator for use in magnetic resonance (MR) flow experiments was evaluated. The accuracy in constant-flow mode was better than 1%. The accuracy in pulsatile-flow mode was found to be dependent on the interconnecting tubing. The short-term and long-term reproducibilities of pulsatile waveforms were less than or equal to 0.4 mL/sec (1 standard deviation). Increased response times due to the lengths of tubing required in MR flow experiments were surmounted by using a modified tubing configuration and precompensated waveforms. Piston reversal was found not to cause major difficulties in MR flow experiments.

A velocity evaluation phantom for colour and pulsed Doppler instruments.

We describe a phantom designed to evaluate the velocity measurements made with colour and pulsed Doppler instruments. Using a belt to translate a large volume of semi-rigid material through the entire Doppler sample volume eliminates many of the problems associated with flow and string phantoms. A servo-motor with feedback circuitry ensures accurate control of the belt velocity with an uncertainty in the mean velocity of 0.14%. The phantom provides velocities with typical variations of 0.07 cm/s. We have demonstrated the usefulness of this phantom by evaluating the linearity and accuracy of three pulsed Doppler instruments over belt velocities ranging from 0 to 80 cm/s. In addition, the measurements show the effects of the wall filter at low belt velocities. Using this phantom, we have quantified the accuracy, linearity and precision of the velocity measurements made by three colour Doppler instruments. The results also show regions where the colour instruments are aliased and where the wall filter dominates.

Computer-controlled positive displacement pump for physiological flow simulation.

A computer-controlled pump for use both in the study of vascular haemodynamics and in the calibration of clinical devices which measure blood flow is designed. The novel design of this pump incorporates two rack-mounted pistons, driven into opposing cylinders by a micro-stepping motor. This approach allows the production of nearly uninterrupted steady flow, as well as a variety of pulsatile waveforms, including waveforms with reverse flow. The capabilities of this pump to produce steady flow from 0.1 to 60 ml s-1, as well as sinusoidal flow and physiological flow, such as that found in the common femoral and common carotid arteries are demonstrated. Cycle-to-cycle reproducibility is very good, with an average variation of 0.1 ml s-1 over thousands of cycles.