Elsevier

Physica Medica

Volume 29, Issue 1, January 2013, Pages 99-110
Physica Medica

Original Paper
Iterative reconstruction methods in two different MDCT scanners: Physical metrics and 4-alternative forced-choice detectability experiments – A phantom approach

https://doi.org/10.1016/j.ejmp.2011.12.004Get rights and content

Abstract

This paper characterizes and evaluates the potential of three commercial CT iterative reconstruction methods (ASIR™, VEO™ and iDose4 ()) for dose reduction and image quality improvement. We measured CT number accuracy, standard deviation (SD), noise power spectrum (NPS) and modulation transfer function (MTF) metrics on Catphan phantom images while five human observers performed four-alternative forced-choice (4AFC) experiments to assess the detectability of low- and high-contrast objects embedded in two pediatric phantoms. Results show that 40% and 100% ASIR as well as iDose4 levels 3 and 6 do not affect CT number and strongly decrease image noise with relative SD constant in a large range of dose. However, while ASIR produces a shift of the NPS curve apex, less change is observed with iDose4 with respect to FBP methods. With second-generation iterative reconstruction VEO, physical metrics are even further improved: SD decreased to 70.4% at 0.5 mGy and spatial resolution improved to 37% (MTF50%). 4AFC experiments show that few improvements in detection task performance are obtained with ASIR and iDose4, whereas VEO makes excellent detections possible even at an ultra-low-dose (0.3 mGy), leading to a potential dose reduction of a factor 3 to 7 (67%–86%). In spite of its longer reconstruction time and the fact that clinical studies are still required to complete these results, VEO clearly confirms the tremendous potential of iterative reconstructions for dose reduction in CT and appears to be an important tool for patient follow-up, especially for pediatric patients where cumulative lifetime dose still remains high.

Introduction

Today, ionizing radiation from computed tomography (CT) scanners represents the greatest per capita medical exposure for the population of industrialized countries [1], [2], [3], [4], [5], [6]. Although this growth is mainly attributed to the increasing number of CT examinations, CT dose per examination is still high and remains an important concern.

Recently, new reconstruction algorithms based on iterative approaches were launched to complete existing strategies already widespread as a way to reduce CT dose during patient examination (tube current modulation, automatic exposure control, high pitch values on dual-source CT, automated kV modulation, etc.) [7], [8], [9], [10]. In 2008, General Electric (Milwaukee, WI, USA) introduced the first CT iterative reconstruction method commercially available for clinical applications, under the name of adaptive statistical iterative reconstruction (ASIR™). ASIR uses a blend of filtered back-projection (FBP) images with iteratively reconstructed images in the raw data domain to reduce the image noise [11], [12], [13], [14]. Two years later, Siemens (Erlangen, Germany) released an iterative reconstruction in image space (IRIS™). Contrary to ASIR, IRIS is based on an iterative reconstruction loop in the image domain to speed up the reconstruction process [15], [16], [17]. In 2011, Philips commercialized the fourth version of its iterative method, which was called iDose4 (). This method attempts to reduce image noise without modifying the noise power spectrum (NPS) in order to keep close to the noise texture of classical FBP images [18], [19]. After noise removal in the raw data domain, an optimal anatomical model is used in the image domain to iteratively eliminate the quantum image noise and to maintain the appearance of full dose image. Toshiba (Tochigi, Japan) has also developed an iterative method, adaptive iterative dose reduction (AIDR™), which is adaptive and automatically calculates the optimized number of iterations [20]. Recently, GE introduced a full iterative reconstruction or model-based iterative reconstruction (MBIR), under the commercial name of VEO™. Unlike its first iterative reconstruction ASIR, VEO is a fully iterative method working in the raw data domain, which takes not only the data statistics into account but also the geometry of the machine itself by considering the voxel volumes of the scanned object, the focal spot size, the active area size of the detector, etc. Moreover, to enhance model precision of the CT scanner, complex mathematical formulations were determined to account for physical effects such as beam hardening, scatter and metal attenuation artifacts [20], [21]. Due to its complexity and specific properties, VEO is currently only designed for acquisitions performed with the Discovery CT750 HD scanner (GE Healthcare, Waukesha, WI, USA) and important reconstruction time is still required to reconstruct CT images.

The number and variety of new reconstruction methods introduces the need of investigating their dose and image quality relationships. Therefore, the purpose of this work was to characterize three commercial iterative reconstructions, ASIR, VEO and iDose4, through standard physical metrics (CT number accuracy, standard deviation (SD), modulation transfer function (MTF) and noise power spectrum) measured on Catphan phantom images. To go a step beyond physical measurements, five human observers performed four-alternative forced-choice (4AFC) experiments under the signal-known-exactly (SKE) paradigm to assess the detectability of low- and high-contrast objects embedded in two pediatric phantoms. Simple contrast and signal-difference-to-noise ratio (SDNR) metrics were also calculated and compared to low-contrast 4AFC results.

Section snippets

Data acquisition and image reconstruction

Images were acquired on two multi-detector CT (MDCT) scanners: (A) a 128-slice Discovery CT750 HD scanner and (B) a 128-slice Ingenuity CT (Philips Healthcare, Cleveland, OH, USA). Both systems were equipped with 64-row detector arrays able to supply up to 40 mm collimation with a standard configuration of 64 × 0.625 mm. Standard physical metrics were measured from Catphan 600 phantom (Phantom Laboratory, Salem, NY, USA) images scanned at 80 and 120 kVp. Because dose reduction is a main concern

Results

The normalized weighted CTDI (nCTDIw) measured at 80 kVp and 120 kVp in the 32 cm CTDI test phantom and at 80 kVp in the 16 cm phantom are presented in Table 2, Table 3. A good agreement (<14%) between console displays and measured values was obtained for both CT units.

Discussion

The purpose of this work was to characterize and compare three iterative reconstruction methods available on two commercial CT units using physical metrics and 4AFC detectability experiments. Because protocol parameters were slightly different (due to CT unit specificities), results were analyzed only when compared to the conventional FBP reconstruction method within the same unit.

For unit A, the second-generation iterative reconstruction VEO showed a large reduction of image noise combined

Conclusion

Based on phantom approaches, our results show that first-generation iterative reconstruction methods ASIR and iDose4 provide similar benefits and mainly act on noise reduction. Clinical studies have confirmed that these methods are able to strongly reduce image noise and allow a significant dose reduction (30–65%) with reconstruction time almost as fast as standard FBP methods [11], [12], [13], [14], [19]. Second-generation iterative reconstruction VEO shows extra noise reduction when compared

Acknowledgments

We would like to thank Professor Olivier Hélénon, chairman of the adult radiology department in the Necker Children’s Hospital of Paris, for his assistance and for providing the machine time required to perform this study as well as Stéphanie Bermon, Ivan Diaz, Eleni Samara and Dimitra Touli, for their participation in this study. This research was supported by the Swiss National Science Foundation grant [# 320030-120382].

References (30)

  • W.A. Kalender et al.

    Technical approaches to the optimization of CT

    Phys Med

    (2008)
  • A. Rose

    Quantum effects in human vision

    Adv Biol Med Phys

    (1957)
  • BFS

    Umweltradioaktivität und Strahlenbelastung im Jahr 2003. Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit (BMU)

    (2003)
  • G. Brix et al.

    Radiation exposure in multi-slice versus single-slice spiral CT: results of a nationwide survey

    Eur Radiol

    (2003)
  • A. Aroua et al.

    Exposure of the Swiss population by radiodiagnostics: 2003 review

    Health Phys

    (2007)
  • C. Shrimpton et al.

    National survey of doses from CT in the UK: 2003

    Br J Radiol

    (2006)
  • F.A. Mettler et al.

    Medical radiation exposure in the U.S. in 2006: preliminary results

    Health Phys

    (2008)
  • UNSCEAR. Sources and effects of ionizing radiation. United Nations Scientific Committee on the Effects of Atomic...
  • C.H. McCollough et al.

    CT dose reduction and dose management tools: overview of available options

    Radiographics

    (2006)
  • M.L.D. Gunn et al.

    State of the art: technologies for computed tomography dose reduction

    Emerg Radiol

    (2009)
  • S. Ulzheimer et al.

    Computed tomography—patient dose and dose reduction technologies

    Health Phys

    (2011)
  • A.K. Hara et al.

    Iterative reconstruction technique for reducing body radiation dose at CT: feasibility study

    AJR

    (2009)
  • A.C. Silva et al.

    Innovations in CT dose reduction strategy: application of the adaptive statistical iterative reconstruction algorithm

    AJR

    (2010)
  • D. Marin et al.

    Low-tube-voltage, high-tube-current multidetector abdominal CT: improved image quality and decreased radiation dose with adaptive statistical iterative reconstruction algorithm – initial clinical experience

    Radiology

    (2010)
  • F.A. Miéville et al.

    Paediatric cardiac CT examinations: impact of the iterative reconstruction method ASIR on image quality – preliminary findings

    Ped Radio

    (2011 Jun 30)
  • Cited by (175)

    View all citing articles on Scopus
    View full text