Impact of noise reduction on radiation dose reduction potential of virtual monochromatic spectral images: Comparison of phantom images with conventional 120 kVp images using deep learning image reconstruction and hybrid iterative reconstruction.

PURPOSE: To assess the effects of deep learning image reconstruction (DLIR) and hybrid iterative reconstruction (HIR) on the image quality of virtual monochromatic spectral (VMS) images and to investigate the dose reduction potential of the VMS and conventional 120 kVp images. METHODS: A cylindrical phantom simulating an adult abdomen was used. The contrast was set to 60 (medium) and 300 (high) Hounsfield units. CT acquisitions were performed at three dose levels: 12, 9, and 6 mGy. Images were reconstructed via filtered back projection (FBP), DLIR, and HIR. The noise power spectrum (NPS) and task transfer function (TTF) were measured, and the system performance (SP) function was calculated (TTF2/NPS). RESULTS: The noise magnitudes at low spatial frequencies with DLIR and HIR were lower than that with FBP by 45.6% and 24.4%, respectively. Compared to the FBP results, the TTF values at 50% with DLIR at medium and high contrast changed by -13.2% and +25.3% with the VMS images and -2.0% and +9.3% with the 120 kVp images, respectively. In the VMS and 120 kVp images, compared to the SP values of 12 mGy FBP images, SP values of 6 mGy DLIR images decreased at medium contrast and increased at high contrast. CONCLUSIONS: DLIR achieved better noise reduction than HIR. The spatial resolution of VMS-DLIR varied significantly depending on the contrast. The image quality of VMS-DLIR and 120 kVp-DLIR potentially decrease in medium contrast tasks and increase in high contrast tasks with 50% dose reduction.

Optimization of pediatric FDG-PET/CT examinations based on physical indicators using the SiPM-PET/CT system.

OBJECTIVE: This study aimed to investigate the appropriate Silicon photomultiplier -PET/CT acquisition and image reconstruction conditions for each age group. METHODS: The original phantom was developed to reflect the thickness and width of the torso in each age group (neonates, 1-year-olds, 5-year-olds, 10-year-olds, 15-year-olds, and adults). The ratio of hot spheres to background radioactivity was 4:1, and the radioactivity concentration was adjusted according to the Japanese consensus guidelines for appropriate implementation of pediatric nuclear medicine examinations. We evaluated the root mean square error (RMSE) as an assessment/function of the standardized uptake value of each hot sphere, the background variability (N10 mm), the % contrast of the hot sphere (QH, 10 mm/N10 mm), and the noise equivalent counts to determine the optimal reconstruction parameters and the appropriate acquisition time. RESULTS: The minimum RMSE was obtained by setting the half-width of the Gaussian filter to 0-2 mm for iteration 1 or 2 and to 2-4 mm for iteration 3 or more. The acquisition times that satisfied the image quality equivalent to 120 s acquisitions in the adult phantoms were 30 s in the neonatal and 1-year-old phantoms, 60 s in the 5- and 10-year-old phantoms, and 75 s in the 15-year-old phantoms. CONCLUSION: This study demonstrated that good PET images could be obtained with short acquisition times when the examination is performed under appropriate reconstruction conditions.

Assessment of the image quality of virtual monochromatic spectral computed tomography images: a phantom study considering object contrast, radiation dose, and frequency characteristics.

Fast kilovoltage (kVp)-switching technology cannot obtain conventional 120 kVp images; thus, 70 keV virtual monochromatic spectral computed tomography (CT) images (VMSI) are generally used. The contrast-to-noise ratio (CNR) is used to evaluate the image quality of VMSI; however, CNR does not include frequency characteristics. The present study aimed to investigate the evaluation methods of VMSI considering frequency characteristics by comparing the image quality of 70 keV VMSI with that of conventional 120 kVp images. The evaluated object contrasts were 70 and 300 Hounsfield units (HU). Scans used two radiation dose levels: low (LD) and standard (SD). The volume CT dose index of LD and SD was 4.8- and 12 mGy, respectively. Images were reconstructed by filtered back projection, evaluating CNR, noise power spectrum (NPS), task transfer function (TTF), and system performance (SP) function calculated as TTF2/ NPS. The total NPS values (spatial frequency range: 0.2 ~ 0.4 mm-1) of 70 keV VMSI were higher than those of 120 kVp images. The spatial frequency TTF values that reached 10% (f10%) of the 70 keV VMSI changed based on object contrast. For the low-contrast condition, a lower f10% was observed with 70 keV VMSI. The CNR of 70 keV VMSI was comparable to that of 120 kVp images in low- and high-contrast conditions. However, for 70 keV VMSI, SP of low-contrast was low, and SP of high-contrast was high, compared with those of 120 kVp images. This study suggested that only CNR was not sufficient to evaluate the image quality of VMSI; thus, evaluation methods considering frequency characteristics should be used.

Spatial resolution compensation by adjusting the reconstruction kernels for iterative reconstruction images of computed tomography.

PURPOSE: Hybrid iterative reconstruction (IR) is useful to reduce noise in computed tomography (CT) images. However, it often decreases the spatial resolution. The ability of high spatial resolution kernels (harder kernels) to compensate for the decrease in the spatial resolution of hybrid IRs was investigated. METHODS: An elliptic cylindrical phantom simulating an adult abdomen was used. Two types of rod-shaped objects with ~330 and ~130 HU were inserted to simulate contrasts of arteries in CT angiography. Two multi-slice CT systems were used to scan the phantoms with 120 kVp and scan doses of 20 and 10 mGy. The task transfer functions (TTFs) were measured from the circular edges of the rod images. The noise power spectrum (NPS) was measured from the images of the water-only section. The CT images were reconstructed using a filtered back projection (FBP) with baseline kernels and two levels of hybrid IRs with harder kernels. The profiles of the clinical images across the aortic dissection flaps were measured to evaluate actual spatial resolutions. RESULTS: The TTF degradation of each hybrid IR was recovered by the harder kernels, whereas the noise reduction effect was retained, for both the 20 and 10 mGy. The profiles of the dissection flaps for the FBP were maintained by using the harder kernels. Even with the best combination of hybrid IR and harder kernel, the noise level at 10 mGy was not reduced to the level of FBP at 20 mGy, suggesting no capability of a 50% dose reduction while maintaining noise.

[Evaluation of Low-contrast Detectability of Iterative Reconstruction Algorithm in High-speed CT Technology for Head].

PURPOSE: The purpose of this study was to assess the feasibility of high-speed CT technology for head without deterioration of low-contrast detectability using the brain LCD (Canon Medical Systems) of iterative reconstruction. METHODS: System performance (SP) function analysis, low-contrast object specific contrast-to-noise ratio (CNRLO) analysis, and visual evaluation using Scheffe's paired comparison were performed. Additionally, analysis of the correlation of CNRLO and visual scores was performed. SP was performed with the self-made phantom. CNRLO was calculated with the catphan 504 phantom (CTP 515). Visual evaluation was performed using the brain phantom which simulated such as cerebral infarction and investigated on a fivepoint scale. All images were acquired with pitch factor of 0.61 (low pitch) and 1.40 (high pitch). All images were reconstructed with filtered back projection (FBP), brain LCD standard (LCD STD) and strong (LCD STR). RESULTS: SP of brain LCD improved compared with FBP. CNRLO of FBP decreased in high pitch compared with low pitch. CNRLO of brain LCD images acquired by low- and high pitch were improved compared with FBP. Visual scores denoted similar trends to that of CNRLO and there was high correlation with CNRLO. CONCLUSION: It was suggested that using brain LCD can achieve the high speed CT technology for head without deterioration of low-contrast detectability.

Quantitative and qualitative evaluation of hybrid iterative reconstruction, with and without noise power spectrum models: A phantom study.

The purpose of this phantom study was to investigate the feasibility of dose reduction with hybrid iterative reconstruction, with and without a noise power spectrum (NPS) model, using both quantitative and qualitative evaluations. Standard dose (SD), three-quarter dose (TQD), and half-dose (HD) of radiation were used. Images were reconstructed with filtered back projection (FBP), adaptive iterative dose reduction 3D (AIDR 3D) (MILD, STR), and AIDR 3D enhanced (eAIDR 3D) (eMILD, eSTR). An NPS analysis, task-based modulation transfer function (MTFtask ) analysis, and comparisons of low-contrast detectability and image texture were performed. Although the eAIDR 3D had a higher NPS value in the high-frequency range and improved image texture and resolution as compared with AIDR 3D at the same radiation dose and iteration levels, it yielded higher noise than AIDR 3D. Additionally, although there was no statistically significant difference between SD-FBP and the TQD series in the comparison of the mean area under the curve (AUC), the mean AUC was statistically significantly different between SD-FBP and the HD series. NPS values in the high-frequency range, 10% MTFtask values, low-contrast detectability, and image textures of TQD-eMILD were comparable to those of SD-FBP. Our findings suggested that using eMILD can reduce the radiation dose by 25%, while potentially maintaining diagnostic performance, spatial resolution, and image texture; this could support selecting the appropriate protocol in a clinical setting.

Evaluation of the CT Parameters to Suppress Renal Cysts Pseudoenhancement Effect: Influence of the Virtual Monochromatic Spectral Images, the Model-based Iterative Reconstruction Algorithm and the Aperture Size in Phantom Model.

PURPOSE: The purpose of this study was to evaluate the effect of the virtual monochromatic spectral images (VMSI) and the model-based iterative reconstruction (MBIR) images, to evaluate the influence of the aperture size (40- and 20-mm beam) on renal pseudoenhancement (PE) compared with the filtered back projection (FBP) images. METHODS: The renal compartment-CT phantom was filled with iodinated contrast material diluted to the attenuation of 180 Hounsfield units (HU) at 120 kV. The water-filled spherical structures, which simulate cyst, were inserted into the renal compartment. Those diameters were 7, 15 and 25 mm. These were scanned by conventional mode (helical scan, 120 kV-FBP) and dual energy mode. 70 keV-VMSI were reconstructed from the dual energy mode, and MBIR images were reconstructed from conventional mode at 40- and 20-mm aperture. Additionally, the phantom was scanned using non-helical mode with 20-mm aperture, and FBP images were reconstructed. The CT value of the PE for cyst areas was measured for these images. RESULTS: The CT values of the cysts were 20.0-14.3 HU on the FBP images, 12.8-12.7 HU on the 70 keV-VMSI (PE-inhibition ratio was 36.0-11.2%) and 16.2-14.0 HU on the MBIR images (19.0-2.1%), respectively, at 40-mm aperture. The PE-inhibition ratio scanned by 20-mm aperture was improved by 28.0% with FBP, 32.8% with 70 keV-VMSI and 29.6% with MBIR compared with 40-mm aperture. One of the FBP images with non-helical mode was 11.6 HU. CONCLUSIONS: The best CT technique to minimize PE was the combination of 70 keV-VMSI and 20-mm aperture.