Skip to main content

Model-based iterative reconstruction technique for radiation dose reduction in chest CT: comparison with the adaptive statistical iterative reconstruction technique



To prospectively evaluate dose reduction and image quality characteristics of chest CT reconstructed with model-based iterative reconstruction (MBIR) compared with adaptive statistical iterative reconstruction (ASIR).


One hundred patients underwent reference-dose and low-dose unenhanced chest CT with 64-row multidetector CT. Images were reconstructed with 50 % ASIR-filtered back projection blending (ASIR50) for reference-dose CT, and with ASIR50 and MBIR for low-dose CT. Two radiologists assessed the images in a blinded manner for subjective image noise, artefacts and diagnostic acceptability. Objective image noise was measured in the lung parenchyma. Data were analysed using the sign test and pair-wise Student’s t-test.


Compared with reference-dose CT, there was a 79.0 % decrease in dose–length product with low-dose CT. Low-dose MBIR images had significantly lower objective image noise (16.93 ± 3.00) than low-dose ASIR (49.24 ± 9.11, P < 0.01) and reference-dose ASIR images (24.93 ± 4.65, P < 0.01). Low-dose MBIR images were all diagnostically acceptable. Unique features of low-dose MBIR images included motion artefacts and pixellated blotchy appearances, which did not adversely affect diagnostic acceptability.


Diagnostically acceptable chest CT images acquired with nearly 80 % less radiation can be obtained using MBIR. MBIR shows greater potential than ASIR for providing diagnostically acceptable low-dose CT images without severely compromising image quality.

Key Points

Model-based iterative reconstruction (MBIR) creates high-quality low-dose CT images.

MBIR significantly improves image noise and artefacts over adaptive statistical iterative techniques.

MBIR shows greater potential than ASIR for diagnostically acceptable low-dose CT.

The prolonged processing time of MBIR may currently limit its routine use in clinical practice.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3



Model-based iterative reconstruction


Adaptive statistical iterative reconstruction


Filtered back projection


Modulation transfer function


Effective dose


  1. Brenner DJ, Hall EJ (2007) Computed tomography—an increasing source of radiation exposure. N Engl J Med 357:2277–2284

    Article  PubMed  CAS  Google Scholar 

  2. Schauer DA, Linton OW (2009) NCRP report no. 160, ionizing radiation exposure of the population of the United States, medical exposure—are we doing less with more, and is there a role for health physicists? Health Phys 97:1–5

    Article  PubMed  CAS  Google Scholar 

  3. Amis ES Jr, Butler PF, Applegate KE et al (2007) American College of Radiology white paper on radiation dose in medicine. J Am Coll Radiol 4:272–284

    Article  PubMed  Google Scholar 

  4. Mettler FA Jr, Bhargavan M, Faulkner K et al (2009) Radiologic and nuclear medicine studies in the United States and worldwide: Frequency, radiation dose, and comparison with other radiation sources – 1950–2007. Radiology 253:520–531

    Article  PubMed  Google Scholar 

  5. Kalra MK, Maher MM, Toth TL et al (2004) Strategies for CT radiation dose optimization. Radiology 230:619–628

    Article  PubMed  Google Scholar 

  6. Heyer CM, Mohr PS, Lemburg SP, Peters SA, Nicolas V (2007) Image quality and radiation exposure at pulmonary CT angiography with 100- or 120-kVp protocol: Prospective randomized study. Radiology 245:577–583

    Article  PubMed  Google Scholar 

  7. Diel J, Perlmutter S, Venkataramanan N, Mueller R, Lane MJ, Katz DS (2000) Unenhanced helical CT using increased pitch for suspected renal colic: An effective technique for radiation dose reduction? J Comput Assist Tomogr 24:795–801

    Article  PubMed  CAS  Google Scholar 

  8. Kalra MK, Maher MM, Sahani DV et al (2003) Low-dose CT of the abdomen: Evaluation of image improvement with use of noise reduction filters pilot study. Radiology 228:251–256

    Article  PubMed  Google Scholar 

  9. Hara AK, Paden RG, Silva AC, Kujak JL, Lawder HJ, Pavlicek W (2009) Iterative reconstruction technique for reducing body radiation dose at CT: Feasibility study. AJR Am J Roentgenol 193:764–771

    Article  PubMed  Google Scholar 

  10. Flicek KT, Hara AK, Silva AC, Wu Q, Peter MB, Johnson CD (2010) Reducing the radiation dose for CT colonography using adaptive statistical iterative reconstruction: A pilot study. AJR Am J Roentgenol 195:126–131

    Article  PubMed  Google Scholar 

  11. Leipsic J, Labounty TM, Heilbron B et al (2010) Adaptive statistical iterative reconstruction: Assessment of image noise and image quality in coronary CT angiography. AJR Am J Roentgenol 195:649–654

    Article  PubMed  Google Scholar 

  12. Leipsic J, Nguyen G, Brown J, Sin D, Mayo JR (2010) A prospective evaluation of dose reduction and image quality in chest CT using adaptive statistical iterative reconstruction. AJR Am J Roentgenol 195:1095–1099

    Article  PubMed  Google Scholar 

  13. Prakash P, Kalra MK, Digumarthy SR et al (2010) Radiation dose reduction with chest computed tomography using adaptive statistical iterative reconstruction technique: Initial experience. J Comput Assist Tomogr 34:40–45

    Article  PubMed  Google Scholar 

  14. Prakash P, Kalra MK, Kambadakone AK et al (2010) Reducing abdominal CT radiation dose with adaptive statistical iterative reconstruction technique. Invest Radiol 45:202–210

    Article  PubMed  Google Scholar 

  15. Sagara Y, Hara AK, Pavlicek W, Silva AC, Paden RG, Wu Q (2010) Abdominal CT: Comparison of low-dose CT with adaptive statistical iterative reconstruction and routine-dose CT with filtered back projection in 53 patients. AJR Am J Roentgenol 195:713–719

    Article  PubMed  Google Scholar 

  16. Singh S, Kalra MK, Gilman MD et al (2011) Adaptive statistical iterative reconstruction technique for radiation dose reduction in chest CT: A pilot study. Radiology 259:565–573

    Article  PubMed  Google Scholar 

  17. Singh S, Kalra MK, Hsieh J et al (2010) Abdominal CT: Comparison of adaptive statistical iterative and filtered back projection reconstruction techniques. Radiology 257:373–383

    Article  PubMed  Google Scholar 

  18. Cornfeld D, Israel G, Detroy E, Bokhari J, Mojibian H (2011) Impact of adaptive statistical iterative reconstruction (ASIR) on radiation dose and image quality in aortic dissection studies: A qualitative and quantitative analysis. AJR Am J Roentgenol 196:W336–W340

    Article  PubMed  Google Scholar 

  19. Thibault JB, Sauer KD, Bouman CA, Hsieh J (2007) A three-dimensional statistical approach to improved image quality for multislice helical CT. Med Phys 34:4526–4544

    Article  PubMed  Google Scholar 

  20. Yu Z, Thibault JB, Bouman CA, Sauer KD, Hsieh J (2011) Fast model-based X-ray CT reconstruction using spatially nonhomogeneous ICD optimization. IEEE Trans Image Process 20:161–175

    Article  PubMed  Google Scholar 

  21. Strub WM, Weiss KL, Sun D (2007) Hybrid reconstruction kernel: Optimized chest CT. AJR Am J Roentgenol 189:W115–W116

    Article  PubMed  Google Scholar 

  22. Weiss KL, Cornelius RS, Greeley AL et al (2011) Hybrid convolution kernel: Optimized CT of the head, neck, and spine. AJR Am J Roentgenol 196:403–406

    Article  PubMed  Google Scholar 

  23. Prakash P, Kalra MK, Ackman JB et al (2010) Diffuse lung disease: CT of the chest with adaptive statistical iterative reconstruction technique. Radiology 256:261–269

    Article  PubMed  Google Scholar 

  24. National Lung Screening Trial Research Team, Aberle DR, Berg CD et al (2011) The National Lung Screening Trial: overview and study design. Radiology 258:243–253

    Article  PubMed  Google Scholar 

  25. Shrimpton PC, Hillier MC, Lewis MA, Dunn M (2006) National survey of doses from CT in the UK: 2003. Br J Radiol 79:968–980

    Article  PubMed  CAS  Google Scholar 

  26. EUR 16262. European guidelines on quality criteria for computed tomography. Accessed 10 February 2012

Download references


We gratefully acknowledge Kosuke Sasaki, M.S., and Koji Segawa, R.T., for their technical support and assistance in data acquisition.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Masaki Katsura.


Appendix 1

ASIR and MBIR are new iterative reconstruction (IR) algorithms. Unlike the conventional FBP, which is based on simpler mathematical assumptions of the tomographic imaging system, IR generates a set of synthesised projections by accurately modelling the data collection process in CT. The model incorporates statistical system information (including photon statistics and electronic noise in the data acquisition system) and details of the system optics (including the size of each detector cell, dimensions of the focal spot, and the shape and size of each image voxel). The synthesised image is mathematically compared and corrected with the actual measurement in order to adjust estimation of the object’s image. The technique then iterates this comparison and correction step in order to achieve close proximity between actual and measured projections. Inconsistencies in the projection measurement due to limited photon statistics and electronic noise are corrected with multiple iterations. These data-processing steps help to improve image quality from the noise and resolution perspectives, but prolong the reconstruction duration compared with FBP because of the intensive computations particularly required for incorporating system optics information.

The ASIR technique models just the photons and electronic noise statistics that primarily affect image noise, which are not as computationally intensive or time-consuming. This enables near real-time display of images at the time of imaging. ASIR also differs from other IR techniques in that the vendor provides a blending tool to blend the FBP with the ASIR images. This is accomplished by reconstruction of CT raw data with both FBP and ASIR techniques and then performing a weighted summation of each data set for the final reconstructed images. Prior phantom and clinical studies have already shown that ASIR provides diagnostically acceptable images with a reduction in image noise for low-radiation dose CT compared with the FBP algorithm [918].

The MBIR technique, on the other hand, is a pure IR technique that does not involve blending with FBP images, and is mathematically more complex and accurate than ASIR. MBIR not only incorporates modelling of photon and noise statistics like ASIR, it also involves modelling of system optics. This is unlike ASIR, which uses an idealised set of system optics (as does FBP), resulting in similar data utilisation per image. MBIR analyses the x-ray beam at the focal spot, then as it passes through the patient’s body and again as it strikes the detector. The algorithm weights each data point so that noisy projections have less influence on the final results, and this allows more accurate image reconstruction. Phantom experiments have shown that MBIR has the potential to further reduce image noise, improve spatial resolution and thereby allow further dose reduction without compromising image quality [19]. With incorporation of system optics information and therefore a more accurate account of voxel and focal spot size and geometry, one can expect improvements in spatial resolution [19, 20]. Because MBIR is a complicated algorithm, using multiple iterations and multiple models, the reconstruction time is significantly longer than FBP as well as the other IR techniques, even with dedicated state-of-the-art parallel processors. The reconstruction time in the present study was about 1 h per case, although the exact time was not recorded as it was not a feature of the application software.

Appendix 2

The rationale for the NI setting in the present study is as follows. As the specified slice thickness for a given NI setting decreases by a factor of x, maintenance of the same radiation dose requires an increase in NI by a factor of 1 divided by the square root of x. According to the previous radiology literature [13, 23], the NI of 15.75 at a slice thickness of 2.5 mm has been used for reference-dose chest CT. Multiplying 15.75 by the square root of 4 (= 2.5/0.625) results in 31.5. Theoretically, the fixed NI of 31.5 at 0.625 mm should achieve the same radiation dose for reference-dose chest CT from the previous radiology literature. As for the NI setting for low-dose CT, we referred to the average ED for low-dose CT described in the National Lung Screening Trial (NLST, 1.5 mSv) [24], which is about one fifth of the radiation dose for reference-dose chest CT described in the previous radiology literature [13, 23]. As the radiation dose decreases by 1/y, the image noise increases by the square root of y. Multiplying the NI for reference-dose CT in the present study (= 31.5) by the square root of 5 results in an NI of 70.44.

Appendix 3

Subjective image noise was defined as overall graininess or mottle in the lung parenchyma, and was assessed in the lung window setting on a five-point scale (1 = no or only minimal image noise, 2 = less than average image noise, 3 = average image noise, 4 = more than average or substantial image noise that may interfere with diagnostic decision-making in less than half of the lung parenchyma, and 5 = more than average or substantial image noise that may interfere with diagnostic decision-making in more than half of the lung parenchyma). Artefacts were graded on a three-point scale (1 = artefacts unapparent or only minimally recognisable, 2 = artefacts recognised but not interfering with diagnostic decision-making, and 3 = substantial artefacts recognised affecting diagnostic decision-making). The following artefacts were assessed: streak artefacts, motion artefacts due to heart wall motion and blotchy pixellated appearance at the tissue interface (Fig. 2b). Critical reproduction of visually sharp anatomical structures was assessed. The following anatomical structures were evaluated: pulmonary fissures; secondary pulmonary lobular structures such as interlobular arteries; large- and medium-sized pulmonary vessels; small pulmonary vessels; large- and medium-sized bronchi; small bronchi; the pleuromediastinal border; the border between the pleura and the thoracic wall; the thoracic aorta; anterior mediastinal structures including thymic residue; the trachea and main bronchi; paratracheal tissue; the carina and lymph node area; and the oesophagus. Diagnostic acceptability was assessed on a four-point scale (1 = fully acceptable, 2 = probably acceptable, 3 = deemed acceptable only for limited clinical conditions, and 4 = unacceptable).

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Katsura, M., Matsuda, I., Akahane, M. et al. Model-based iterative reconstruction technique for radiation dose reduction in chest CT: comparison with the adaptive statistical iterative reconstruction technique. Eur Radiol 22, 1613–1623 (2012).

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


  • Model-based iterative reconstruction
  • Adaptive statistical iterative reconstruction
  • Radiation dose reduction
  • Image noise
  • Spatial resolution