Iterative Reconstruction Technique for Reducing Body Radiation Dose at CT: Feasibility Study

Department of Diagnostic Radiology, Mayo Clinic Arizona, 13400 E Shea Blvd., Scottsdale, AZ 85259, USA.
American Journal of Roentgenology (Impact Factor: 2.73). 09/2009; 193(3):764-71. DOI: 10.2214/AJR.09.2397
Source: PubMed


The purpose of this study was to evaluate the image noise, low-contrast resolution, image quality, and spatial resolution of adaptive statistical iterative reconstruction in low-dose body CT.
Adaptive statistical iterative reconstruction was used to scan the American College of Radiology phantom at the American College of Radiology reference value and at one-half that value (12.5 mGy). Test objects in low- and high-contrast and uniformity modules were evaluated. Low-dose CT with adaptive statistical iterative reconstruction was then tested on 12 patients (seven men, five women; average age, 67.5 years) who had previously undergone routine-dose CT. Two radiologists blinded to scanning technique evaluated images of the same patients obtained with routine-dose CT and low-dose CT with and without adaptive statistical iterative reconstruction. Image noise, low-contrast resolution, image quality, and spatial resolution were graded on a scale of 1 (best) to 4 (worst). Quantitative noise measurements were made on clinical images.
In the phantom, low- and high-contrast and uniformity assessments showed no significant difference between routine-dose imaging and low-dose CT with adaptive statistical iterative reconstruction. In patients, low-dose CT with adaptive statistical iterative reconstruction was associated with CT dose index reductions of 32-65% compared with routine imaging and had the least noise both quantitatively and qualitatively (p < 0.05). Low-dose CT with adaptive statistical iterative reconstruction and routine-dose CT had identical results for low-contrast resolution and nearly identical results for overall image quality (grade 2.1-2.2). Spatial resolution was better with routine-dose CT (p = 0.004).
These preliminary results support body CT dose index reductions of 32-65% when adaptive statistical iterative reconstruction is used. Studies with larger statistical samples are needed to confirm these findings.

  • Source
    • "Previous reports on IR in CT have demonstrated its efficacy in detecting urinary stone disease with reduced noise and radiation dose. Dose reduction by use of the IR technique does not seem to affect stone characteristics such as stone size, volume, SSD, and HU, which are essential parameters in making treatment decisions [22,23,24,25]. Stone size in CT is a major factor that can affect treatment options. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Purpose To evaluate the efficacy of low-dose computed tomography (LDCT) for detecting urinary stones with the use of an iterative reconstruction technique for reducing radiation dose and image noise. Materials and Methods A total of 101 stones from 69 patients who underwent both conventional nonenhanced computed tomography (CCT) and LDCT were analyzed. Interpretations were made of the two scans according to stone characteristics (size, volume, location, Hounsfield unit [HU], and skin-to-stone distance [SSD]) and radiation dose by dose-length product (DLP), effective dose (ED), and image noise. Diagnostic performance for detecting urinary stones was assessed by statistical evaluation. Results No statistical differences were found in stone characteristics between the two scans. The average DLP and ED were 384.60±132.15 mGy and 5.77±1.98 mSv in CCT and 90.08±31.80 mGy and 1.34±0.48 mSv in LDCT, respectively. The dose reduction rate of LDCT was nearly 77% for both DLP and ED (p<0.01). The mean objective noise (standard deviation) from three different areas was 23.0±2.5 in CCT and 29.2±3.1 in LDCT with a significant difference (p<0.05); the slight increase was 21.2%. For stones located throughout the kidney and ureter, the sensitivity and specificity of LDCT remained 96.0% and 100%, with positive and negative predictive values of 100% and 96.2%, respectively. Conclusions LDCT showed significant radiation reduction while maintaining high image quality. It is an attractive option in the diagnosis of urinary stones.
    Full-text · Article · Sep 2014 · Korean journal of urology
  • Source
    • "The main goal of IR algorithms is to combine low image noise with little radiation exposure. ASIR technique accurately rebuilds images obtained from the FBP algorithm [18] by trying to lower the image noise using complex matrix algebra [19]. For this data reconstruction, the ASIR algorithm transforms each pixel value (y) that is measured in the FBP protocol to a new estimate of the pixel value (y ). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Purpose To assess ASIR (adaptive statistical iterative reconstruction) technique regarding dose reduction and its impact on image quality in evaluation CTs of potential kidney donors. Materials and Methods Between May and November 2013, a prospective study of 53 assumingly healthy potential kidney donors was conducted. The subjects underwent abdominal evaluation CT prior to the planned explantation of a kidney and were randomly divided into 2 groups: Group A was examined with an ASIR 40 protocol (n = 26), group B (n = 27) was examined using a standard FBP (filtered back projection) protocol. Image quality was assessed both quantitatively (by obtaining attenuation values in different organ regions and calculating SNR and CNRs) and qualitatively (by two observers who evaluated image quality using a 5-point scale system). Applied dose was analysed as CTDIvol (mGy), total DLP (mGy x cm) and effective dose (mSv). Results Applied dose in group A was about 26% lower than in group B (p < 0.05). Between both groups, dose determining parameters such as scan length and patients‘body diameter showed no significant difference. SNR (signal-to-noise ratio) was significantly higher in group A (p < 0.05). CNRs (contrast-to-noise ratios) for different tissues were not significantly different. Observer rated image quality showed no significant difference. Conclusion ASIR can contribute to a relevant dose reduction without any loss of image quality in CT scans for evaluating potential kidney donors.
    Full-text · Article · Aug 2014 · European Journal of Radiology
  • Source
    • "Teknik lain untuk mereduksi dosis adalah penggunaan metode iterative reconstruction (IR) [26] [27] [28] [29] [30] [31] [32]. Dengan teknik ini dosis dapat diturunkan hingga 60% [2]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: ABSTRAK TEKNIK REKONSTRUKSI ITERATIF UNTUK DATA PROYEKSI RENGGANG (SPARSE PROJECTION DATA) SEBAGAI UPAYA UNTUK MEREDUKSI DOSIS CT SCAN. Telah dilakukan upaya untuk mereduksi dosis pada CT Scan menggunakan data proyeksi renggang (sparse projection data) dan menggunakan rekonstruksi metode iteratif, yaitu dengan algoritma maximum likelihood expectation-maximization (ML-EM). Rekonstruksi citra dilakukan dengan software Matlab. Obyek yang digunakan adalah fantom Shepp-Logan dengan ukuran piksel 512 x 512. Berkas yang digunakan adalah berkas paralel dari sudut 1-180 derajat, dengan interval sudut divariasi dari 1 derajat hingga 10 derajat. Hasil rekosntruksi dengan metode ML-EM dibandingkan dengan metode filtered back-projection (FBP) dengan jenis filter Ram-Lak, sebagai acuan. Besaran yang digunakan untuk menentukan kualitas citra adalah mean absolute error (MAE). Diperoleh bahwa nilai MAE metode FBP untuk interval sudut 1 derajat sebesar 0,0191. Untuk mencapai nilai MAE 0,0191 pada metode ML-EM pada interval sudut 1 derajat dibutuhkan iterasi sebanyak 30 kali. Untuk interval sudut sampai 5 derajat, MAE 0,0191 hanya dibutuhkan sedikit penambahan jumlah iterasi, yaitu tidak lebih dari 3 iterasi. Namun, pada interval sudut melebihi 3 derajat, mulai terjadi streak artifact, meskipun nilai MAE sama yaitu sebesar 0,0191. Dengan demikian, metode ini dapat mereduksi dosis CT Scan hingga 66,7%. Kata kunci: Iterative reconstruction (IR), algoritma MLEM, CT dose reduction, sparse projection data ABSTRACT ITERATIVE RECONSTRUCTION TECHNIQUE FOR DISPERSE DATA PROJECTIONS (SPARSE PROJECTION DATA) AS A REDUCTION EFFORTS FOR CT SCAN DOSE. Efforts have been made to reduce the dose in CT scans using disperse projection data (sparse projection data) and using iterative reconstruction methods, namely the algorithm of maximum likelihood expectation-maximization (ML-EM Image reconstruction is done with Matlab. The object used is Shepp-Logan phantom with a pixel size of 512 x 512. The file that is used is a parallel beam of angle 1-180 degrees , with intervals varied angle of 1 degree to 10 degrees. Results rekosntruksi with ML-EM method than the method of filtered back-projection (FBP) with a Ram-Lak filter types , as a reference. Quantity that is used to determine the quality of the image is the mean absolute error (MAE). Provided that the value of MAE FBP method for interval angle of 1 degree at 0.0191. To achieve MAE value 0.0191 on the ML-EM method at intervals of 1 degree angle of iterations needed 30 times. For the interval angle to 5 degrees , MAE 0.0191 only a slight increase in the number of iterations required , ie no more than three iterations. However , the angle interval exceeds 3 degrees , started happening streak artifact , although the same MAE value that is equal to 0.0191. Thus , this method can reduce the dose CT scans up to 66.7 % .
    Full-text · Article · Jun 2014
Show more