CT Iterative Reconstruction
Computed Tomography Iterative Reconstruction · Also known as: MBIR, ASIR, IR-CT, statistical reconstruction
CT Iterative Reconstruction (IR) is a computational technique that reconstructs tomographic images from raw X-ray projection data by iteratively refining an estimate of tissue attenuation until it matches the measured projections. Developed from algebraic reconstruction techniques pioneered by Gordon in 1974, iterative reconstruction has revolutionized clinical CT by enabling high-quality images at reduced radiation dose. Variants such as Adaptive Statistical Iterative Reconstruction (ASIR) and Model-Based Iterative Reconstruction (MBIR) are now standard on modern CT scanners.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Iterative reconstruction is indicated in all CT studies, especially those where dose reduction is prioritized (pediatric imaging, pregnant patients, high-risk individuals, or longitudinal monitoring). It is particularly valuable in low-dose screening CT (lung cancer, coronary calcium), trauma imaging where fast acquisition is needed, and in patients where conventional FBP images are unacceptably noisy. IR is less critical for high-dose CT (adult chest with thick slices) where noise is minimal with FBP, but IR provides incremental benefit. IR requires iterative reconstruction hardware and software, available on all modern scanners (GE, Siemens, Philips, Toshiba/Canon).
Strengths & limitations
- Radiation dose reduction: IR achieves diagnostic image quality at 30-80% lower dose than FBP, reducing lifetime cancer risk and enabling broader screening
- Noise reduction: eliminates the noise amplification inherent to FBP, improving low-contrast lesion detectability especially in soft tissue and lung parenchyma
- Artifact reduction: minimizes beam hardening, streak, and motion artifacts by enforcing consistency between forward-projected guesses and measured data
- Edge preservation: IR maintains anatomical edges and subtle contrast transitions better than FBP, aiding diagnostic interpretation
- Quantitative accuracy: reduced noise and artifacts improve quantitative measurements (density, perfusion, nodule volume), enabling more reliable longitudinal comparisons
- Computational cost: IR is 10-100 times slower than FBP, requiring dedicated GPU hardware and increasing total scan-to-diagnosis time (though clinical workstations can process in parallel)
- Algorithm specificity: different vendor algorithms (ASIR, MBIR, iDose, AIDR, FIRST) are not directly comparable; switching vendors or algorithms affects noise texture and appearance
- Learning curve for radiologists: IR images have different noise characteristics (texture is less granular, more blotchy) than FBP; radiologists must learn to recognize normal IR appearance to avoid false-positive diagnoses
- Limited high-noise benefit: IR is most effective in moderate-noise regimes; at very high noise (very low dose), IR cannot restore spatial resolution lost to Poisson statistics
- Incomplete standardization: no universally agreed-upon metrics for IR strength or quality; clinical settings are often empirically optimized without formal guidelines
Frequently asked
What is the difference between ASIR and MBIR?
ASIR (Adaptive Statistical Iterative Reconstruction) is a hybrid method that blends IR with FBP, typically applying IR to only part of the reconstruction (20-100% blend). It is faster than pure IR. MBIR (Model-Based IR) is full iterative reconstruction that incorporates detailed physical models of the scanner and noise, typically requiring more iterations. MBIR achieves higher dose reduction (30-40%) but is slower. ASIR is more widely adopted for speed; MBIR is preferred when maximal dose reduction is essential.
Does iterative reconstruction eliminate the need for dose reduction protocols?
No. IR enables safe dose reduction but does not replace optimized acquisition protocols. Optimized scan parameters (mA, pitch, collimation, tube potential) should be set first; IR then minimizes noise from the lower dose. Combining aggressive dose reduction with IR can push dose too low, reintroducing noise and artifacts despite IR.
How does IR change the appearance of images compared to FBP?
IR images have a different noise texture: less granular and grainy (FBP) and more blotchy or smoothed (IR). Subtle artifacts may appear different. Edge clarity is improved. Radiologists trained exclusively on FBP must relearn normal anatomy in IR images to avoid misinterpreting noise texture as pathology. Side-by-side comparison during transition is valuable.
Can IR artifacts be confused with disease?
Yes. In some IR algorithms, especially MBIR with high strength, artificial hypoattenuation can appear within tumors or lesions, mimicking necrosis or cystic change. Radiologists should correlate with other sequences (ultrasound, MRI) when suspicious. Known artifacts (blooming at metallic implants) may appear different with IR than FBP.
Is iterative reconstruction available on all modern CT scanners?
Essentially yes. All major vendors (GE, Siemens, Philips, Toshiba/Canon) offer iterative reconstruction. Implementation and naming vary: GE ASIR/MBIR, Siemens ADMM/SAFIRE, Philips iDose, Toshiba AIDR. Performance differs; clinical sites often validate locally and select reconstruction parameters specific to their scanner and anatomy of interest.
Sources
- Gordon, R., Bender, R., Herman, G. T. (1974). Algebraic reconstruction techniques (ART) for three-dimensional electron microscopy and X-ray photography. Journal of Theoretical Biology, 29(3), 471-481. link ↗
- Yu, L., Leng, S., McCollough, C. H. (2012). Iterative reconstruction in medical imaging. Journal of Medical Imaging, 1(3), 033506. link ↗
- Singh, S., Kalra, M. K., Hsieh, J., et al. (2010). Abdominal CT: comparison of adaptive statistical iterative and filtered back projection reconstruction techniques. Radiology, 257(2), 373-383. DOI: 10.1148/radiol.10092212 ↗
How to cite this page
ScholarGate. (2026, June 3). Computed Tomography Iterative Reconstruction. ScholarGate. https://scholargate.app/en/medical-imaging/ct-iterative-reconstruction
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- DEXAMedical Imaging↔ compare
- DTI TractographyMedical Imaging↔ compare
- PET Kinetic ModelingMedical Imaging↔ compare
- Quantitative Susceptibility MappingMedical Imaging↔ compare
- RadiomicsMedical Imaging↔ compare