Comparison of adaptive statistical iterative and filtered back projection reconstruction techniques in brain CT

https://doi.org/10.1016/j.ejrad.2011.12.041Get rights and content

Abstract

Purpose

To compare image quality and visualization of normal structures and lesions in brain computed tomography (CT) with adaptive statistical iterative reconstruction (ASIR) and filtered back projection (FBP) reconstruction techniques in different X-ray tube current–time products.

Materials and methods

In this IRB-approved prospective study, forty patients (nineteen men, twenty-one women; mean age 69.5 ± 11.2 years) received brain scan at different tube current–time products (300 and 200 mAs) in 64-section multi-detector CT (GE, Discovery CT750 HD). Images were reconstructed with FBP and four levels of ASIR-FBP blending. Two radiologists (please note that our hospital is renowned for its geriatric medicine department, and these two radiologists are more experienced in chronic cerebral vascular disease than in neoplastic disease, so this research did not contain cerebral tumors but as a discussion) assessed all the reconstructed images for visibility of normal structures, lesion conspicuity, image contrast and diagnostic confidence in a blinded and randomized manner. Volume CT dose index (CTDIvol) and dose-length product (DLP) were recorded. All the data were analyzed by using SPSS 13.0 statistical analysis software.

Results

There was no statistically significant difference between the image qualities at 200 mAs with 50% ASIR blending technique and 300 mAs with FBP technique (p > .05). While between the image qualities at 200 mAs with FBP and 300 mAs with FBP technique a statistically significant difference (p < .05) was found.

Conclusion

ASIR provided same image quality and diagnostic ability in brain imaging with greater than 30% dose reduction compared with FBP reconstruction technique.

Introduction

There has been several image reconstruction algorithms developed since the introduction of computed tomography (CT) technique. Until recently, the Filtered Back Projection (FBP) reconstruction technique was considered as the standard CT imaging reconstruction algorithm due to its short reconstruction times that despite some disadvantages most noticeably the image noise. This reconstruction technique is based on direct inversion of Radon transform [1]. Although the use of ramp filter in FBP removes blurring from the back projection step [2], the limited number of projection sets introduces streak artifacts [3] in the image reconstructions. A different algorithm, namely, the iterative reconstruction algorithm was investigated in order to reduce image noise among other things. The Algebraic Reconstruction Technique (ART) was one of the first iterative reconstruction algorithm used in the commercial clinical CT scanner [1]. However, earlier versions of iterative reconstruction algorithms were complex and required a lot of computational power, and were not widely used in clinical applications.

Adaptive statistical iterative reconstruction (ASIR) algorithm which was introduced in 2009 has caught more and more radiologists’ attention as a newly developed technique with its ability to produce real time images in clinical practice. In contrast to FBP, ASIR takes into account precise modeling of the X-ray photon statistics and electronic noise, all of which are less accurate in FBP [4]. To decrease reconstruction time, ASIR utilizes the information contained in the FBP-reconstruction image as an initial ‘building block’ in the reconstruction process [5]. By partially correcting for the fluctuations in projection measurement due to limited photon statistics, ASIR enables a time-efficient reduction in the pixel variance that is statistically unlikely to be representative of anatomic features, with essentially no trade-off in spatial resolution. In a recent study, Hara et al. thought ASIR was similar to the FBP in terms of spatial resolution [6].

It was reported that ASIR can better reduce image noise, radiation dose and improve diagnostic confidence than FBP reconstruction technique in abdominal, lung, coronary artery and colon CT studies [4], [7], [8], [9], but no articles were found in brain CT about ASIR. For this purpose, we conducted this study to investigate whether ASIR algorithm can improve the image quality and reduce radiation dose in brain CT scanning.

Section snippets

Patient selection

This prospective clinical study was compliant with the guidelines of the Health Insurance Portability and Accountability Act (HIPAA). All patients received written informed consent.

The study included forty patients who underwent brain CT examinations between April 2011 and June 2011. The forty patients were divided into four groups according to their ages (group A, 50–59; group B, 60–69; group C, 70–79; group D, 80–). Brain CT examinations were ordered by attending physicians for patients with

Objective image quality measurements

Statistical results of the objective image quality measurements were summarized in Table 1. There were no statistically significant differences between the MSD of the four age groups with same reconstruction technique (p > .05). The MSD decreased with increasing percentage of ASIR at each level of tube current–time product (mAs); the difference between FBP and 30% ASIR blending at each level of mAs had no statistical significance (p > .05). In addition, there was no statistically significant

Discussion

Since the first commercial clinical computer tomographic (CT) scanner was born, it has played a more and more important role in clinical practice, especially nowadays with its substantial improvements over the recent decades. But the average medical radiation effective dose (ED) to the U.S. population in 2006 was estimated at approximately 3.0 mSv, an increase of 600% in a single generation [12]. The growing use of imaging procedures in the United States has raised concerns about exposure to

Conflict of interest

We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the manuscript.

References (18)

  • S. Vandenberghe et al.

    Iterative reconstruction algorithms in nuclear medicine

    Comput Med Imaging Graphics

    (2001)
  • G. Wang et al.

    An outlook on X-ray research and development

    Med Phys

    (2008)
  • R.H. Huesman

    Effects of a finite number of projection angles and finite lateral sampling of projections on the propagation of statistical errors in transverse section reconstruction

    Phys Med Biol

    (1977)
  • S. Singh et al.

    Abdominal CT: comparison of adaptive statistical iterative and filtered back projection reconstruction techniques

    Radiology

    (2010)
  • A.C. Silva et al.

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

    Am J Roentgenol

    (2010)
  • A.K. Hara et al.

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

    Am J Roentgenol

    (2009)
  • P. Prakash et al.

    Diffuse lung disease: CT of the chest with adaptive statistical iterative reconstruction technique

    Radiology

    (2010)
  • J. Leipsic et al.

    Adaptive statistical iterative reconstruction: assessment of image noise and image quality in coronary CT angiography

    Am J Roentgenol

    (2010)
  • K.T. Flicek et al.

    Reducing the radiation dose for CT colonography using adaptive statistical iterative reconstruction: a pilot study

    Am J Roentgenol

    (2010)
There are more references available in the full text version of this article.

Cited by (52)

  • Neuroimaging for the Primary Care Provider: A Review of Modalities, Indications, and Pitfalls

    2021, Pediatric Clinics of North America
    Citation Excerpt :

    In the past decade, CT technology has rapidly advanced, which has allowed images to be obtained with ever-decreasing ionizing radiation dose and compensating for patient motion. Recent advances in CT techniques and technology have led to a reduction in ionizing radiation dose with noncontrast head CT between 22% and 43%.7–9 There have been other recent advances that reduce motion artifacts.

  • The Acceptability of Iterative Reconstruction Algorithms in Head CT: An Assessment of Sinogram Affirmed Iterative Reconstruction (SAFIRE) vs. Filtered Back Projection (FBP) Using Phantoms

    2017, Journal of Medical Imaging and Radiation Sciences
    Citation Excerpt :

    Although all images reviewed in this study had the same values for objective image quality measures, the quality of the institutional reference standard protocol images was judged inferior than those of optimized protocols owing to perceived levels of subjective image sharpness and the ability to detect low-contrast lesions. Importantly, consistent with other research [16, 18, 19, 23, 24], the findings indicate a potential maximum relative dose-reduction limit of up to 24% for head CT examinations with SAFIRE reconstruction. However, other institutions prefer to practice on the side of caution and report only 15% dose reduction for CT head [25].

  • Benefits of sinogram-affirmed iterative reconstruction in 0.4 mSv ultra-low-dose CT of the upper abdomen following transarterial chemoembolisation: Comparison to low-dose and standard-dose CT and filtered back projection technique

    2016, Clinical Radiology
    Citation Excerpt :

    Iterative reconstruction promises to improve image quality even while lowering the radiation dose. This has already been shown in CT of the head,1–5 paranasal sinuses,6 cervical spine,7 coronary arteries,8–17 chest,18–24 and abdomen.25–33 In the latter case, SIR showed promising results in a phantom study on low-contrast objects, as well for the evaluation of hepatic metastasis and hepatocellular carcinoma; however, the extent of dose reduction in CT after transarterial chemoembolisation (TACE) of liver tumours or metastases is currently unknown.

View all citing articles on Scopus

Sources of funding: Supported by Science and Technology Commission Foundation of Shanghai (No. 10411952600).

View full text