The feasibility of a piecewise‐linear dynamic bowtie filter

Abstract

Purpose: The prepatient attenuator (or "bowtie filter") in CT is used to modulate the flux as a function of fan angle of the x-ray beam incident on the patient. Traditional, static bowtie filters are tailored only for very generic scans and for the average patient. The authors propose a design for a dynamic bowtie that can produce a time-dependent piecewise-linear attenuation profile. This dynamic bowtie may reduce dynamic range, dose or scatter, but in this work they focus on its ability to reduce dynamic range, which may be particularly important for systems employing photon-counting detectors. Methods: The dynamic bowtie is composed of a set of triangular wedges. Each wedge is independently moved in order to produce a time-dependent piecewise-linear attenuation profile. Simulations of the bowtie are conducted to estimate the dynamic range reduction in six clinical datasets. The control of the dynamic bowtie is determined by solving a convex optimization problem, and the dose is estimated using Monte Carlo techniques. Beam hardening artifacts are also simulated. Results:The dynamic range is reduced by factors ranging from 2.4 to 27 depending on the part of the body studied. With a dynamic range minimization objective, the dose to the patient can be reduced from 6% to 33% while maintaining peak image noise. Further reduction in dose may be possible with a specific dose reduction objective. Beam hardening artifacts are suppressed with a two-pass algorithm. Conclusions: A dynamic bowtie producing a time-dependent, piecewise-linear attenuation profile is possible and can be used to modulate the flux of the scanner to the imaging task. Initial simulations show a large reduction in dynamic range. Several other applications are possible.

Task-driven optimization of CT tube current modulation and regularization in model-based iterative reconstruction

“…Multi-location optimization using TCM alone will therefore only be able to consider the most challenging location, or improve average or weighted average performance over all locations. A technology more suited to multi-location optimization is (Stayman et al 2016, Szczykutowicz and Mistretta 2014, Hsieh and Pelc 2013, Bartolac et al 2011) which generalized the TCM concept to provide varying spatial modulations of the x-ray beam as a function of rotation angle have been developed. Initial studies generalizing the TCM strategies presented here are underway to permit greater freedom in providing locally optimal fluence-field modulation to multiple locations in the image (Gang et al 2016).…”

“…The optimal modulation pattern is dependent on the local spatial resolution and noise properties and is therefore location-dependent. Prescribing optimal fluence patterns for many locations requires additional flexibility in data acquisition – e.g, dynamic fluence field modulation (Stayman et al 2016, Hsieh and Pelc 2013, Szczykutowicz and Mistretta 2014, Bartolac et al 2011, Gang et al 2016) – which is the subject of ongoing investigation.…”

Vision 20/20: Single photon counting x‐ray detectors in medical imaging

“…Another design has a set of triangular wedges and each wedge independently moves longitudinally. 68 A third design has a hollow ellipse which rotates in the direction opposite to the gantry rotation. …”

Joint optimization of fluence field modulation and regularization for multi-task objectives

“…Various hardware solutions 1–3 have demonstrated potential for achieving FFM on diagnostic CT scanners. What FFM patterns should be delivered, however, remains an open question.…”