[Validation of Optimal Imaging Conditions for Coronary Computed Tomography Angiography Using High-definition Mode and Deep Learning Image Reconstruction Algorithm].

PURPOSE: To verify the optimal imaging conditions for coronary computed tomography angiography (CCTA) examinations when using high-definition (HD) mode and deep learning image reconstruction (DLIR) in combination. METHOD: A chest phantom and an in-house phantom using 3D printer were scanned with a 256-row detector CT scanner. The scan parameters were as follows - acquisition mode: ON (HD mode) and OFF (normal resolution [NR] mode), rotation time: 0.28 s/rotation, beam coverage width: 160 mm, and the radiation dose was adjusted based on CT-AEC. Image reconstruction was performed using ASiR-V (Hybrid-IR), TrueFidelity Image (DLIR), and HD-Standard (HD mode) and Standard (NR mode) reconstruction kernels. The task-based transfer function (TTF) and noise power spectrum (NPS) were measured for image evaluation, and the detectability index (d') was calculated. Visual evaluation was also performed on an in-house coronary phantom. RESULT: The in-plane TTF was better for the HD mode than for the NR mode, while the z-axis TTF was lower for DLIR than for Hybrid-IR. The NPS values in the high-frequency region were higher for the HD mode compared to those for the NR mode, and the NPS was lower for DLIR than for Hybrid-IR. The combination of HD mode and DLIR showed the best value for in-plane d', whereas the combination of NR mode and DLIR showed the best value for z-axis d'. In the visual evaluation, the combination of NR mode and DLIR showed the best values from a noise index of 45 HU. CONCLUSION: The optimal combination of HD mode and DLIR depends on the image noise level, and the combination of NR mode and DLIR was the best imaging condition under noisy conditions.

Task-based assessment of resolution properties of CT images with a new index using deep convolutional neural network.

In this study, we propose a method for obtaining a new index to evaluate the resolution properties of computed tomography (CT) images in a task-based manner. This method applies a deep convolutional neural network (DCNN) machine learning system trained on CT images with known modulation transfer function (MTF) values to output an index representing the resolution properties of the input CT image [i.e., the resolution property index (RPI)]. Sample CT images were obtained for training and testing of the DCNN by scanning the American Radiological Society phantom. Subsequently, the images were reconstructed using a filtered back projection algorithm with different reconstruction kernels. The circular edge method was used to measure the MTF values, which were used as teacher information for the DCNN. The resolution properties of the sample CT images used to train the DCNN were created by intentionally varying the field of view (FOV). Four FOV settings were considered. The results of adapting this method to the filtered back projection (FBP) and hybrid iterative reconstruction (h-IR) images indicated highly correlated values with the MTF10% in both cases. Furthermore, we demonstrated that the RPIs could be estimated in the same manner under the same imaging conditions and reconstruction kernels, even for other CT systems, where the DCNN was trained on CT systems produced by the same manufacturer. In conclusion, the RPI, which is a new index that represents the resolution property using the proposed method, can be used to evaluate the resolution of a CT system in a task-based manner.

Evaluation of the second-generation whole-heart motion correction algorithm (SSF2) used to demonstrate the aortic annulus on cardiac CT.

The main purpose of pre-transcatheter aortic valve implantation (TAVI) cardiac computed tomography (CT) for patients with severe aortic stenosis is aortic annulus measurements. However, motion artifacts present a technical challenge because they can reduce the measurement accuracy of the aortic annulus. Therefore, we applied the recently developed second-generation whole-heart motion correction algorithm (SnapShot Freeze 2.0, SSF2) to pre-TAVI cardiac CT and investigated its clinical utility by stratified analysis of the patient's heart rate during scanning. We found that SSF2 reconstruction significantly reduced aortic annulus motion artifacts and improved the image quality and measurement accuracy compared to standard reconstruction, especially in patients with high heart rate or a 40% R-R interval (systolic phase). SSF2 may contribute to improving the measurement accuracy of the aortic annulus.

Performance of Ultra-High-Resolution Computed Tomography in Super High-Resolution Mode at the Routine Radiation Dose: Phantom Study.

OBJECTIVE: Using a chest phantom, we compared the image quality of ultra-high-resolution computed tomography (U-HRCT) images acquired in super high-resolution (SHR) and normal resolution (NR) mode and at the routine radiation dose. The detector size was 0.25 and 0.5 mm, respectively. METHODS: A chest phantom was scanned on a U-HRCT scanner. The scan parameters were tube voltage 120 kV and volume CT dose index 13.0 mGy, the routine radiation dose for conventional scans. The rotation time was 0.5 s/rot, the number of matrices was 512 in NR and 1024 in SHR mode. For physical evaluation, the modulation transfer function was measured on the spherical simulated nodule, and the noise power spectrum on the cylindrical water phantom. A CT value profile curve was created using an in-house simulated bronchial phantom. For visual evaluation, 3 radiologists and 3 radiology technologists evaluated overall image quality using a 4-grade scale (grade 1, poor; and grade 4, excellent). RESULTS: The 10% of modulation transfer function was 13.5 lp/cm in NR and 14.9 lp/cm in SHR mode ( P <0.01). ƒ peak was 5.6 lp/cm in NR and 8.8 lp/cm in SHR mode ( P <0.01), and the peak of noise power spectrum shifted. On the profile curves, the CT value at the edge changed in NR but not in SHR mode. The overall image quality was grade 3.0 ± 0.7 in SHR and grade 2.0 ± 0.7 in NR mode ( P <0.01). CONCLUSIONS: The image quality of SHR mode with U-HRCT was superior to that of NR mode at the routine radiation dose.

[Proposal for Reduction Measures of Eye Lens Exposure Based on Actual Exposure Management in Radiation-exposed Medical Staff].

PURPOSE: To investigate the actual condition of the crystalline lens equivalent dose and effective dose according to the type of job and the type of duties in a medical institution. We also sought to clarify effective exposure reduction strategies. METHODS: Equivalent crystalline lens doses, effective doses, job type, and duties for 8656 persons · year were obtained from 17 medical facilities. We analyzed the relationship between the effective dose and the crystalline lens equivalent dose in uniform exposure control and non-uniform exposure control conditions. Exposure data were obtained for 13 unique job types and duties. RESULTS: The ratio of the lens equivalent dose to the effective dose of non-uniform exposure managers was 2 to 6 times and varied depending on the occupation. The percentage of persons whose annual lens equivalent dose exceeded 20 mSv was 4.75% for medical doctors, 1.17% for nurses, and 0.24% for radiological technologists. Highly exposed tasks included doctors in cardiology and gastroenterology performing angiography and endoscopy, nurses in endoscopy, and radiological technologists in radiography and CT examinations. CONCLUSION: Thorough unequal exposure control for operations with high crystalline lens exposure, radiation protection education, and effective use of proper personal protective equipment such as the use of radiation protection glasses may reduce lens exposure levels.

Deep learning-based reconstruction in ultra-high-resolution computed tomography: Can image noise caused by high definition detector and the miniaturization of matrix element size be improved?

PURPOSE: This study aimed to assess the noise characteristics of ultra-high-resolution computed tomography (UHRCT) with deep learning-based reconstruction (DLR). METHODS: Two different diameters of water phantom were scanned with three different resolution acquisition modes. Images were reconstructed by filtered back projection (FBP), hybrid iterative reconstruction (hybrid-IR), and DLR. Image noise analysis was performed with noise magnitude, peak frequency (fp) of the noise power spectrum (NPS), and the square root of the area under the curve (√AUCNPS) for the NPS curve. RESULTS: The noise magnitude was up to 3.30 times higher for the FBP acquired in SHR mode than that for the NR mode. The fp values of the FBP were 0.20-0.21, 0.34-0.36, and 0.34-0.37 cycles/mm for normal resolution (NR), high resolution (HR), and super high resolution (SHR) mode, respectively. The fp of hybrid-IR was 0.16-0.19, 0.21-0.26, and 0.23-0.26 cycles/mm for NR, HR, and SHR mode, respectively. The fp of DLR was 0.21-0.32 and 0.22-0.33 cycles/mm for HR and SHR mode, respectively. √AUCNPS showed that the highest value in FBP images of the SHR mode was up to 1.89 times that of the NR mode. DLR in the HR and SHR modes showed high noise reduction while suppressing fp shift with respect to FBP. CONCLUSIONS: The new DLR algorithm could be a solution to the noise increase due to the high-definition detector elements and the small reconstruction matrix element size.

[Organ-based Tube-current Modulation Applied on Different MDCT Scanners: Reduction in the Radiation Dose to the Eye Lens at Head CT].

PURPOSE: Organ-based tube current modulation (OB-TCM) techniques, which are provided by three vendors, reduces the radiation dose to the lens of the eyes by decreasing the tube current, when the X-ray tube passes over the anterior surface of critical organs. However, the characteristics of dose modulation of these techniques are different. The purpose of this study was to understand the performance characteristics of OB-TCM technique of each computed tomography (CT) vendor at head CT. METHODS: We used three CT scanners (SOMATOM Definition Flash; Siemens Healthcare, Revolution CT; GE Healthcare, and Aquilion ONE Genesis Edition; Canon Medical Systems). We measured the radiation dose to the lens surface as evaluation of radiation dose reduction and measured the image noise as index of image quality. We measured the radiation dose rate in the air for analysis of the characteristics of dose modulation in each OB-TCM. RESULTS: When applying OB-TCM, the radiation doses for the lens surface were decreased by 28%, 22%, and 25% for Siemens, GE, and Canon CT scanners, respectively, and the image noise level was increased by 5.6%, 8.5%, and 15.1% for Siemens, GE, and Canon CT scanners, respectively. The characteristics of dose modulation in each OB-TCM were also confirmed by measured the radiation dose rate. CONCLUSION: We confirmed that each OB-TCM has different influence on image quality and radiation doses for lens surface, due to the different characteristics of dose modulation for each CT vendor.

Angular dependence of shielding effect of radiation protective eyewear for radiation protection of crystalline lens.

Radiation protective (RP) eyewear effectively protects crystalline lenses from radiation exposure. A drawback of RP eyewear is the angular dependence of the shielding effect, which results from the design of the eyewear. In this study, 21 models of RP eyewear with different designs and lead equivalences were assessed. Each piece of RP eyewear was hung on a Styrofoam phantom that imitated the head, and a 0.125-cc ionization chamber dosimeter was placed at the position of the crystalline lens. The differences in angular dependence of the shielding effect were evaluated by changing the irradiation angle, and parameters that improved the angular dependence of the shielding effect-sufficient lead equivalence, large coverage design, and minimum gap between the crystalline lens and the RP eyewear-were identified. Thus, the findings highlight the importance of selecting RP eyewear according to the angular distribution and the nature of radiation exposure in the workplace for radiation workers.

Overranging and overbeaming measurement in area detector computed tomography: A method for simultaneous measurement in volume helical acquisition.

PURPOSE: We propose a novel method to assess overbeaming and overranging, as well as the effect of reducing longitudinal exposure range, by using a dynamic z-collimator in area detector computed tomography. METHODS AND MATERIALS: A 500-mm diameter cylindrical imaging plate was exposed by helical scanning in a dark room. The beam collimation of the helical acquisitions was set at 32 and 80 mm. Overbeaming and overranging with the dynamic z-collimator were measured. RESULTS: The actual beam widths were approximately 39 and 88 mm at 32 and 80 mm collimation, respectively, and were relatively reduced owing to increased beam collimation. Overranging was 27.0 and 48.2 mm with a pitch of 0.83 and 1.49 at 32 mm collimation and 72.5 and 83.1 mm with a pitch of 0.87 and 0.99 at 80 mm collimation. The dynamic z-collimator relatively reduced the overranging by 17.3% and 17.1% for the 32 and 80 mm collimation, respectively. CONCLUSION: We devised a method to simultaneously measure overbeaming and overranging with only one helical acquisition. Although the dynamic z-collimator reduced the overranging by approximately 17%, wider collimation widths and higher pitch settings would increase the exposure dose outside the scan range.

Assessment of longitudinal beam property and contrast uniformity for 256- and 320-row area detector computed tomography scanners in the 160-mm nonhelical volume-acquisition mode.

BACKGROUND: Because the x-ray property of patient longitudinal axis in area detector computed tomography (ADCT) depends on a heel effect, radiation dose and beam quality are not uniform along the long axis of the patient. OBJECTIVE: This study aimed to measure the longitudinal beam properties and contrast uniformity of ADCT scanners in the 160-mm nonhelical volume-acquisition (NVA) mode and provide useful datasets for the radiation dose reduction in ADCT examinations. MATERIALS AND METHODS: Two different types of ADCT scanners were used in this study. To assess the heel effect in 256- and 320-row ADCT scanners, we measured dose profile, half-value layer, and iodine contrast uniformity along longitudinal beam direction. RESULTS: The maximum effective energy difference within a 160-mm x-ray beam is approximately 4 keV. Maximum radiation dose on the anode side of the x-ray tube showed approximately 40%-45% reduction compared with that on the isocenter position; the heel effect properties longitudinally differed throughout the x-ray beam, and the decrease in the radiation dose in 256- and 320-row ADCT scanners was observed on the patient table side and gantry side respectively. The CT numbers of iodinated solutions for 256-row ADCT scanner were independent of the heel effect; nevertheless, the CT numbers of 320-row ADCT scanner tended to increase on the patient table (cathode) side. CONCLUSION: This study reveals that the radiation dose on the anode side of the x-ray tube shows approximately 40%-45% reduction compared with that on the isocenter position, and the heel effect properties for 256- and 320-row ADCT scanners longitudinally differ throughout the x-ray beam. The x-ray tube for individual ADCT scanners is mounted in an opposite direction along the long axis of the patient.

Effective temporal resolution and image quality of volume scanning in 320-row detector CT.

To measure the effective temporal resolution (eTR) and image quality for three reconstruction modes for non-helical volume scanning in area detector CT. Temporal sensitivity profiles (TSPs) were obtained and the full width of the TSP at half maximum was used as an index of the eTR. Image quality was assessed by image noise and the corrected artifact index. The half reconstruction mode had a higher eTR than the full and automatic patient motion collection (APMC) reconstructions. Compared to full reconstruction, the image noise with APMC and half reconstruction were increased by 16% and 35%. The corrected artifact index was lowest with APMC. The square root of full width at tenth maximum of the TSP showed a high coefficient of determination (R2 = 0.934) for image noise. This study revealed the TSPs and eTRs for non-helical volume scanning in area detector CT. A high eTR resulted in higher image noise.

Objective assessment of low-contrast computed tomography images with iterative reconstruction.

OBJECTIVE: This study aims to assess low-contrast image quality using a low-contrast object specific contrast-to-noise ratio (CNRLO) analysis for iterative reconstruction (IR) computed tomography (CT) images. METHODS: A phantom composed of low-contrast rods placed in a uniform material was used in this study. Images were reconstructed using filtered back projection (FBP) and IR (Adaptive Iterative Dose Reduction 3D). Scans were performed at six dose levels: 1.0, 1.8, 3.1, 4.6, 7.1 and 13.3mGy. Objective image quality was assessed by comparing CNRLO with CNR using a human observer test. RESULTS: Compared with FBP, IR yielded increased CNR at the same dose levels. The results of CNRLO and observer tests showed similarities or only marginal differences between FBP and IR at the same dose levels. The coefficient of determination for CNRLO was significantly better (R(2)=0.86) than that of CNR (R(2)=0.47). CONCLUSION: For IR, CNRLO could potentially serve as an objective index reflective of a human observer assessment. The results of CNRLO test indicated that the IR algorithm was not superior to FBP in terms of low-contrast detectability at the same radiation doses.

Temporal resolution measurement of 128-slice dual source and 320-row area detector computed tomography scanners in helical acquisition mode using the impulse method.

PURPOSE: To analyse the temporal resolution (TR) of modern computed tomography (CT) scanners using the impulse method, and assess the actual maximum TR at respective helical acquisition modes. METHODS: To assess the actual TR of helical acquisition modes of a 128-slice dual source CT (DSCT) scanner and a 320-row area detector CT (ADCT) scanner, we assessed the TRs of various acquisition combinations of a pitch factor (P) and gantry rotation time (R). RESULTS: The TR of the helical acquisition modes for the 128-slice DSCT scanner continuously improved with a shorter gantry rotation time and greater pitch factor. However, for the 320-row ADCT scanner, the TR with a pitch factor of <1.0 was almost equal to the gantry rotation time, whereas with pitch factor of >1.0, it was approximately one half of the gantry rotation time. The maximum TR values of single- and dual-source helical acquisition modes for the 128-slice DSCT scanner were 0.138 (R/P=0.285/1.5) and 0.074s (R/P=0.285/3.2), and the maximum TR values of the 64×0.5- and 160×0.5-mm detector configurations of the helical acquisition modes for the 320-row ADCT scanner were 0.120 (R/P=0.275/1.375) and 0.195s (R/P=0.3/0.6), respectively. CONCLUSION: Because the TR of a CT scanner is not accurately depicted in the specifications of the individual scanner, appropriate acquisition conditions should be determined based on the actual TR measurement.

[Effects of differences in head holder on image quality and radiation dose in head CT].

PURPOSE: For emergency or pediatric head CT scans, a simplified pillow made of hard sponge instead of a dedicated head holder may be used if it is difficult to immobilize the head. However, the radiation dose when using a simplified head holder may be increased due to radiation absorption by the patient couch if the automatic exposure control (AEC) system is used. In this phantom study, we compared the radiation dose delivered when using a dedicated and a simplified head holder. MATERIALS AND METHODS: We used a dedicated-type and a pillow-type head holder made of hard sponge (simplified head holder). We placed a 20 cm-diameter cylindrical phantom made of water-equivalent material and an anthropomorphic head phantom in the head holders and then scanned them five times with a 64-detector CT scanner (VCT, GE Healthcare). We performed step-and-shoot and helical scanning with AEC; the noise index was set to 2.8. We measured the radiation dose using fluorescent glass dosimeters in the head phantom and the image noise at five sites in the cylindrical phantom. All values were averaged. RESULTS: With step-and-shoot scans, the mean image noise with the dedicated and the simplified head holder was 3.30 ± 0.05 [SD] and 3.20 ± 0.05, respectively. With helical scans they were 3.00 ± 0.09 and 2.88 ± 0.03, respectively. There was no statistically significant difference (p = 0.02 and 0.04, Student's t-test). The radiation doses with the dedicated and the simplified head holder were 58.6 and 70.4 mGy, respectively, for step-and-shoot scanning and 41.8 and 49.0 mGy, respectively, for helical scanning. The doses were thus significantly higher with the simplified head holder for both step-and-shoot and helical scanning (p < 0.01 and < 0.01). CONCLUSION: We recommend the use of a dedicated head holder for head scanning with AEC since the radiation dose was lower than with the simplified head holder.

Spatial resolution measurement for iterative reconstruction by use of image-averaging techniques in computed tomography.

The purpose of our study was to investigate the validity of a spatial resolution measuring method that uses a combination of a bar-pattern phantom and an image-averaging technique, and to evaluate the spatial resolution property of iterative reconstruction (IR) images with middle-contrast (50 HU) objects. We used computed tomography (CT) images of the bar-pattern phantom reconstructed by the IR technology Adaptive Iterative Dose Reduction 3D (AIDR 3D), which was installed in the multidetector CT system Aquilion ONE (Toshiba Medical Systems, Otawara, Japan). The contrast of the bar-pattern image was set to 50 HU, which is considered to be a middle contrast that requires higher spatial resolution clinically. We employed an image-averaging technique to eliminate the influence of image noise, and we obtained averaged images of the bar-pattern phantom with sufficiently low noise. Modulation transfer functions (MTFs) were measured from the images. The conventional wire method was also used for comparison; in this method, AIDR 3D showed MTF values equivalent to those of filtered back projection. For the middle-contrast condition, the results showed that the MTF of AIDR 3D decreased with the strength of IR processing. Further, the MTF of AIDR 3D decreased with dose reduction. The image-averaging technique used was effective for correct evaluation of the spatial resolution for middle-contrast objects in IR images. The results obtained by our method clarified that the resolution preservation of AIDR 3D was not sufficient for middle-contrast objects.

[Novel method of noise power spectrum measurement for computed tomography images with adaptive iterative reconstruction method].

Adaptive iterative reconstruction techniques (IRs) can decrease image noise in computed tomography (CT) and are expected to contribute to reduction of the radiation dose. To evaluate the performance of IRs, the conventional two-dimensional (2D) noise power spectrum (NPS) is widely used. However, when an IR provides an NPS value drop at all spatial frequency (which is similar to NPS changes by dose increase), the conventional method cannot evaluate the correct noise property because the conventional method does not correspond to the volume data natures of CT images. The purpose of our study was to develop a new method for NPS measurements that can be adapted to IRs. Our method utilized thick multi-planar reconstruction (MPR) images. The thick images are generally made by averaging CT volume data in a direction perpendicular to a MPR plane (e.g. z-direction for axial MPR plane). By using this averaging technique as a cutter for 3D-NPS, we can obtain adequate 2D-extracted NPS (eNPS) from 3D NPS. We applied this method to IR images generated with adaptive iterative dose reduction 3D (AIDR-3D, Toshiba) to investigate the validity of our method. A water phantom with 24 cm-diameters was scanned at 120 kV and 200 mAs with a 320-row CT (Acquilion One, Toshiba). From the results of study, the adequate thickness of MPR images for eNPS was more than 25.0 mm. Our new NPS measurement method utilizing thick MPR images was accurate and effective for evaluating noise reduction effects of IRs.