The Study Of IDose~4Advanced Iteraitve Reconstruciton Technique On Radiation Dose Reduction With Chest Computed Tomography

【Background】With recent advances in computed tomography (CT), the increasing frequency of CTexaminations make the effect radiation dose(ED) an important issue, the focus transitionsfrom only changes scanning parameters to improve the image quality. Research has shownthat30mAs~50mAs is the most appropriate radiation dose for low-dose chest CTexamination, in addition, artifacts will increase in the apex level when radiation isdecreased to20mAs, thereby reducing the detection rate of lung nodules[1]. Research onlow-dose CT is transitioning from a focus on scanning parameters to improving imagereconstruction techniques. Analytical reconstruction (AR) and iterative reconstruction (IR)are the two basic reconstruction techniques. Currently most commercial CT systems usedfiltered back projection (FBP) as the standard image reconstruction algorithm for its fastreconstruction speed and simple building operations[2]. IR algorithm once had been used tothe first CT machine, but was abandoned for its slow reconstruction speed and complexbuilding operations[3].Compared to the FBP technology, IR algorithm can significantlyreduce the dosage while improve the spatial resolution, and selective remove the noise[4].Therefore, with the low dose pressing for the development trend, iterative reconstructiontechnique have to back to clinical applications[5]. The manufacturers have introduceddifferent image reconstruction techniques, such as Siemens IRIS、Toshiba ADIR、GE ASIRand Philips-iDose~4. It can transfer raw data for subsequent off-line reconstruction directlyand all the images were reconstructed to1mm section thicknesses by using IRreconstruction techniques. The purpose of the study was to investigate whether the imagequality could be maintained or even improved in a low-dose non-contrast chest CT usingiDose~4IR reconstruction technique compared with the standard FBP technique in initialexperimental and clinical research.Part ⅠEffect of iDose~4advanced iterative reconstruction on image quality and identificationof ground-glass opacities and partly solid nodules on low-dose CT using chestphantom【Objective】To assess the influence of iDose~4advanced iterative reconstruction (IR) comparedwith filtered back projection (FBP) algorithm on image quality and identification ofground-glass opacities (GGOs) and partly solid nodules on low-dose thin-section CT. 【Material and Methods】A chest CT phantom including simulated GGOs and partly solid nodules was scannedwith four different tube currents: standard control group:120kVp/100mAs (group A),Low-dose protocols:120kVp/50mAs (group B),120kVp/25mAs (group C) and120kVp/15mAs (group D). All the raw data were reconstructed using FBP and iDose~4IR (1-6)reconstruction levels. Quantitative CT numbers and standard deviation (SD) of lung weremeasured on the workstation, followed by visually assessment of identification and imagequality of GGOs and partly solid nodules by two chest radiologists. The standard referenceused100mAs in conjunction with FBP. At the same time, a single radiation dose wascalculated.【Results】1、The ED of the four groups was2.45mSv、1.22mSv、0.61mSv、0.36mSv,the meandose of low-dose group B,C and D was reduced by50.37%,75.1%and85.3%comparedwith group A respectively.2、The CT numbers of lung don’t show significant difference under different tubecurrents and algorithms (all P>0.05). iDose~4IR technique improved the image noisecompared with FBP for the same tube currents. MSD in low-dose groups with FBPalgorithm showed a statistically significant difference compared with100mAs with FBPalgorithm, while no significant differences were found in50mAs iDose~4critical level-1,25mAs iDose~4critical level-4,15mAs iDose~4-6compared with100mAs with FBPalgorithm, and assessment of identification for nodules showed that25mAs reconstructedwith iDose~4-4and15mAs with iDose~4-6was comparable with standard reference(0.4≤kappa<0.75).【Conclusion】Compared with FBP, iDose~4can improve image noise, for phantom study50mAs withiDose~4-1,25mAs with critical level-4, and15mAs with level-4as the suitable low-dosechest computed tomography scanning protocols. Part ⅡRadiation dose and image quality analysis of non-enhanced chest CT after usingiDose4IR technique：the clinical research.【Objective】 To evaluate the image quality and radiation dose of chest computed tomography usingiDose~4IR technique.【Materials and Methods】144patients underwent non-contrast chest CT using a256-slice iCT scanner (PhilipsHealthcare, Cleveland, OH, US), they were randomly assigned into four groups: group A(n=40) standard dose protocol120kvp/100mAs; Low-dose protocols: group B (n=40)120kv/50mAs; group C (n=40)120kvp/25mAs; group D (n=24)120kvp/15mAs; All theraw dates were reconstructed using FBP and iterative reconstruction technique (iDose~4).Group A with filtered back projection (FBP) was a standard control group. Quantitative CTnumbers and standard deviation (MSD), signal-to-noise ratio (SNR), contrast-to-noise ratio(CNR) and the subjective image quality were compared between and within groups. At thesame time, a single radiation dose was calculated.【Results】1、The mean dose of low-dose group B,C and D was reduced by52.38%,75.8%and83.15%compared with group A respectively.2、The CT numbers of lung don’t show significant difference under different tube currentsand algorithms (all P>0.05). iDose~4IR technique improved the image noise, SNR andCNR. No significant differences were found in MSD for the three lung levels between50mAs and100mAs with FBP. MSD in25mAs with iDose~4-4and6were reducedabout34.02%、47.83%compared with FBP algorithms, MSD,SNR and CNR don’tshow significant difference compared with group A. MSD of15mAs with iDose~4-6showed no difference with control group but SNR and CNR were all lower than100mAs with FBP.【Conclusion】25mAs with iDose~4critical level-4can reduce radiation dose by75.8%meanwhileimprove image quality can be used as the suitable low-dose chest computed tomographyscanning protocol.