Effect of radiation dose reduction on image quality in adult head CT with noise-suppressing reconstruction system with a 256 slice MDCT

Abstract

The purpose of our study was to investigate the effect of iterative reconstruction (IR) as a dose reduction system on the image quality (IQ) of the adult head computed tomography (CT) at various low-dose levels, and to identify ways of setting the amount of dose reduction. We performed two noncontrast low-dose (LD) adult head CT protocols modified by lowering the tube current with IR which were decided in the light of a group of phantom studies. Two groups of patients, each 100 underwent noncontrast head CT with LD-I and LD-II, respectively. These groups were compared with 100 consecutive standard dose (STD) adult head CT protocol in terms of quantitative and qualitative IQ. The signal-to-noise ratio (SNR) of the white matter (WM) and gray matter (GM) and contrast-to-noise ratio (CNR) values in the LD groups were higher than the STD group. The differences were statistically significant. When the STD and the LD groups were compared qualitatively, no significant differences were found in overall quality. By selecting the appropriate level of IR 34%, radiation dose reduction in adult head CT can be achieved without compromising IQ.

Figure 1

CIRS phantom: (a) lateral scout; (b) the axial slice scanned at 300 mAs, demonstrating region‐of‐interest placement.

Figure 2

Catphan low‐contrast phantom module (CTP 515) is made of several sets of cylindrical low‐contrast objects: (a) STD protocol (300 mAs, with FBP reconstruction); (b) LD‐I (250 mAs, with iDose⁴2); (c) LD‐II (200 mAs, iDose⁴3). Comparison of low‐contrast images showed comparable visualization of the 7 disc objects, up to 4 mm diameter, at 1% in all of the protocols similar to data published in the literature. $\text{FBP}=\text{filtered backprojection};\text{LD}=\text{low dose};\text{STD}=\text{standard dose}$.

Figure 3

Catphan high‐resolution phantom module (CTP 528) with high‐resolution bar patterns: (a) STD protocol (300 mAs, with FBP reconstruction); (b) LD‐I (250 mAs, with iDose⁴2); (c) LD‐II (200 mAs, iDose⁴3). There were no differences in the identification of these high‐contrast resolution bar patterns between the groups. $\text{LD}=\text{low dose};\text{FBP}=\text{filtered backprojection};\text{STD}=\text{standard dose}$.

Figure 4

CT axial images of various protocols: (a) and (d) STD protocol (300 mAs, with FBP reconstruction); (b) and (e) LD‐I (250 mAs, with iDose⁴2); (c) and (f) LD‐II (200 mAs, iDose⁴3). The overall IQ is similar. $\text{FBP}=\text{filtered backprojection};\text{IQ}=\text{image quality};\text{LD}=\text{low dose};\text{STD}=\text{standard dose}$.