Iterative reconstructions in multiphasic CT imaging of the liver: qualitative and task-based analyses of image quality

Comparison of two versions of a deep learning image reconstruction algorithm on CT image quality and dose reduction: A phantom study

PURPOSE

To compare the impact on CT image quality and dose reduction of two versions of a Deep Learning Image Reconstruction algorithm.

MATERIAL AND METHODS

Acquisitions on the CT ACR 464 phantom were performed at five dose levels (CTDI_vol : 10/7.5/5/2.5/1 mGy) using chest or abdomen pelvis protocol parameters. Raw data were reconstructed using the filtered-back projection (FBP), the enhanced level of AIDR 3D (AIDR 3De), and the three levels of AiCE (Mild, Standard, and Strong) for the two versions (AiCE V8 vs AiCE V10). The noise power spectrum (NPS) and task-based transfer function (TTF) for bone (high-contrast insert) and acrylic (low-contrast insert) inserts were computed. To quantify the changes of noise magnitude and texture, the square root of the area under the NPS curve and the average spatial frequency (f_av ) of the NPS curve were measured. The detectability index (d') was computed to model the detectability of either a large mass in the liver or lung, or a small calcification or high contrast tissue boundaries.

RESULTS

The noise magnitude was lower with both AiCE versions than with AIDR 3De. The noise magnitude was lower with AiCE V10 than with AiCE V8 (-4 ± 6% for Mild, -13 ± 3% for Standard, and -48 ± 0% for Strong levels). f_av and TTF_50% values for both inserts shifted towards higher frequencies with AiCE than with AIDR 3De. Compared to AiCE V08, f_av shifted towards higher frequencies with AiCE V10 (45 ± 4%, 36 ± 3%, and 5 ± 4% for all levels, respectively). The TTF_50% values shifted towards higher frequencies with AiCE V10 as compared with AiCE V8 for both inserts, except for the Strong level for the acrylic insert. Whatever the dose and AiCE levels, d' values were higher with AiCE V10 than with AiCE V8 for the small object/calcification and for the large object/lesion.

CONCLUSION

As compared to AIDR 3De, lower noise magnitude and higher spatial resolution and detectability index were found with both versions of AiCE. As compared to AiCE V8, AiCE V10 reduced noise and improved spatial resolution and detectability without changing the noise texture in a simple geometric phantom, except for the Strong level. AiCE V10 seems to have a greater potential for dose reduction than AiCE V8.

Individually Body Weight-Adapted Contrast Media Application in Computed Tomography Imaging of the Liver at 90 kVp

OBJECTIVES

The aim of the present study was to evaluate the attenuation and image quality (IQ) of a body weight-adapted contrast media (CM) protocol compared with a fixed injection protocol in computed tomography (CT) of the liver at 90 kV.

MATERIALS AND METHODS

One hundred ninety-nine consecutive patients referred for abdominal CT imaging in portal venous phase were included. Group 1 (n = 100) received a fixed CM dose with a total iodine load (TIL) of 33 g I at a flow rate of 3.5 mL/s, resulting in an iodine delivery rate (IDR) of 1.05 g I/s. Group 2 (n = 99) received a body weight-adapted CM protocol with a dosing factor of 0.4 g I/kg with a subsequent TIL adapted to the patients' weight. Injection time of 30 seconds was kept identical for all patients. Therefore, flow rate and IDR changed with different body weight. Patients were divided into 3 weight categories; 70 kg or less, 71 to 85 kg, and 86 kg or greater. Attenuation (HU) in 3 segments of the liver, signal-to-noise ratio, and contrast-to-noise ratio were used to evaluate objective IQ. Subjective IQ was assessed by a 5-point Likert scale. Differences between groups were statistically analyzed (P < 0.05 was considered statistically significant).

RESULTS

No significant differences in baseline characteristics were found between groups. The CM volume and TIL differed significantly between groups (P < 0.01), with mean values in group 1 of 110 mL and 33 g I, and in group 2 of 104.1 ± 21.2 mL and 31.2 ± 6.3 g I, respectively. Flow rate and IDR were not significantly different between groups (P > 0.05). Body weight-adapted protocoling led to more homogeneous enhancement of the liver parenchyma compared with a fixed protocol with a mean enhancement per weight category in group 2 of 126.5 ± 15.8, 128.2 ± 15.3, and 122.7 ± 21.2 HU compared with that in group 1 of 139.9 ± 21.4, 124.6 ± 24.8, and 116.2 ± 17.8 HU, respectively.

CONCLUSIONS

Body weight-adapted CM injection protocols result in more homogeneous enhancement of the liver parenchyma at 90 kV in comparison to a fixed CM volume with comparable objective and subjective IQ, whereas overall CM volume can be safely reduced in more than half of patients.

Improving image quality with model-based iterative reconstruction at quarter of nominal dose in upper abdominal CT

OBJECTIVE

To evaluate the ability of a model-based iterative reconstruction (MBIR) for improving image quality in upper abdominal CT with quarter of the normal dose, in comparison with adaptive statistical iterative reconstruction (ASiR) at normal dose.

METHODS

40 upper abdominal patients were randomly divided into two groups: normal-dose group (n = 20) with tube current modulation for noise index (NI) of 10 HU and 40% ASiR reconstruction; low-dose group (n = 20) with NI = 20  HU in the delay phase and MBIR and 40%ASiR. Images in the delay phase were compared. The CT values and standard deviation (SD) values of the liver, spleen, pancreas, kidney, erector spine and fat were measured. Contrast-noise-ratio (CNR = (CT_tissue-CT _fat)/SD_fat) of each measured organ were calculated and compared with one-way ANOVA among the three reconstruction groups. The subjective image scores of the three groups were assessed blindly by two experienced physicians using a 5-point system and the score consistency was compared by the κ test.

RESULTS

Dose reduction of 75 % was achieved for the low-dose scan. The subjective scores (95 % confidence intervals) of the three groups (NI 10-40 %  ASiR, NI 20-40%  ASiR and NI 20-MBIR) were 4.00 ± 0.79 (3.62-4.37), 3.35 ± 0.58 (3.07-3.62) and 3.90 ± 0.64 (3.60-4.19), respectively with no difference between the NI 10-40%  ASiR and NI20-MBIR groups and good consistency between reviewers (κ = 0.726). MBIR had statistically lower SD values and higher contrast-to-noise ratio values in the liver, spleen, pancreas, kidney and erector spine than NI 10-40%  ASiR and NI 20-40%  ASiR (all p < 0.05).

CONCLUSION

At 75 % dose reduction, MBIR provides similar image quality compared to 40% ASiR at normal-dose.

ADVANCES IN KNOWLEDGE

MBIR provides good image quality at 25 % of the normal dose.