Intra-individual diagnostic image quality and organ-specific-radiation dose comparison between spiral cCT with iterative image reconstruction and z-axis automated tube current modulation and sequential cCT

Deep learning assisted differentiation of hepatocellular carcinoma from focal liver lesions: choice of four-phase and three-phase CT imaging protocol

PURPOSE

To evaluate whether a three-phase dynamic contrast-enhanced CT protocol, when combined with a deep learning model, has similar accuracy in differentiating hepatocellular carcinoma (HCC) from other focal liver lesions (FLLs) compared with a four-phase protocol.

METHODS

Three hundred and forty-two patients (mean age 49.1 ± 10.5 years, range 19-86 years, 65.8% male) scanned with a four-phase CT protocol (precontrast, arterial, portal-venous and delayed phases) were retrospectively enrolled. A total of 449 FLLs were categorized into HCC and non-HCC groups based on the best available reference standard. Three convolutional dense networks (CDNs) with the input of four-phase CT images (model A), three-phase images without portal-venous phase (model B) and three-phase images without precontrast phase (model C) were trained on 80% of lesions and evaluated in the other 20% by receiver operating characteristics (ROC) and confusion matrix analysis. The DeLong test was performed to compare the areas under the ROC curves (AUCs) of A with B, B with C, and A with C.

RESULTS

The diagnostic accuracy in differentiating HCC from other FLLs on test sets was 83.3% for model A, 81.1% for model B and 85.6% for model C, and the AUCs were 0.925, 0.862 and 0.920, respectively. The AUCs of models A and C did not differ significantly (p = 0.765), but the AUCs of models A and B (p = 0.038) and of models B and C (p = 0.028) did.

CONCLUSIONS

When combined with a CDN, a three-phase CT protocol without precontrast showed similar diagnostic accuracy as a four-phase protocol in differentiating HCC from other FLLs, suggesting that the multiphase CT protocol for HCC diagnosis might be optimized by removing the precontrast phase to reduce radiation dose.

Improved Peritoneal Cavity and Abdominal Organ Imaging Using a Biphasic Contrast Agent Protocol and Spectral Photon Counting Computed Tomography K-Edge Imaging

OBJECTIVES

To validate in vitro the capability of a high-spatial-resolution prototype spectral photon-counting computed tomography (SPCCT) scanner to differentiate between 2 contrast agents and to assess in vivo the image quality and the feasibility to image the peritoneal cavity in rats using the 2 contrast agents simultaneously within the vascular and peritoneal compartments.

MATERIALS AND METHODS

The authors performed SPCCT imaging (100 mAs, 120 kVp) with energy bin thresholds set to 30, 51, 64, 72, and 85 keV in vitro on a custom-made polyoxymethylene cylindrical phantom consisting of tubes with dilutions of both contrast agents and in vivo on 2 groups of adult rats using 2 injection protocols. Approval from the institutional animal ethics committee was obtained. One group received macrocylic gadolinium chelate intraperitoneal (IP) and iodine intravenous (IV) injections (protocol A, n = 3), whereas the second group received iodine IP and gadolinium IV (protocol B, n = 3). Helical scans were performed 35 minutes after IP injection and 20 seconds after IV injection. The SPCCT and contrast material images, that is, iodine and gadolinium maps, were reconstructed with a field of view of 160 mm, an isotropic voxel size of 250 μm, and a matrix size of 640 × 640 pixels using a soft reconstruction kernel. The SPCCT images were reconstructed with 2 different spatial resolutions to compare the image quality (sharpness, diagnostic quality, and organ visualization) of SPCCT (250 μm) with single-energy computed tomography (CT) (600 μm). Two radiologists evaluated the peritoneal opacification index in 13 regions (score = 0-3 per region) on each type of image. Concentrations of contrast agents were measured in the organs of interest.

RESULTS

In vitro, the concentration measurements correlated well with the expected concentrations. The linear regressions both had R values of 0.99, slopes of 0.84 and 0.87, and offsets at -0.52 and -0.38 mg/mL for iodine and gadolinium, respectively. In vivo, the SPCCT images were of better diagnostic quality, with increased sharpness compared with the CT-like images (P < 0.0001). Intraperitoneal diffusion was excellent, with similar peritoneal opacification index on SPCCT images and overlay of contrast material maps (P = 1) without a significant difference between protocol A (37.0 ± 1.7) and protocol B (35.3 ± 1.5) (P = 0.34). Only the contrast material maps demonstrated clear visual separation of the contrast agents, allowing specific quantification of the physiological enhancement in the liver, spleen, and kidney and the urinary clearance in the renal pelvis and bladder. Renal excretion of the contrast agents injected IP was observed and was consistent with blood diffusion.

CONCLUSIONS

Spectral photon-counting CT can be used to perform a complete peritoneal dual-contrast protocol, enabling a good assessment of the peritoneal cavity and abdominal organs in rats.

Radiation dose and image quality in intraoperative CT (iCT) angiography of the brain with stereotactic head frames

OBJECTIVES

Intraoperative CT (iCT) angiography of the brain with stereotactic frames is an integral part of navigated neurosurgery. Validated data regarding radiation dose and image quality in these special examinations are not available. We therefore investigated two iCT protocols in this IRB-approved study.

METHODS

Retrospective analysis of patients, who received a cerebral stereotactic iCT angiography on a 128 slice CT scanner between February 2016 and December 2017. In group A, automated tube current modulation (ATCM; reference value 410 mAs) and automated tube voltage selection (reference value 120 kV) were enabled, and only examinations with a selected voltage of 120 kV were included. In group B, fixed parameters were applied (300 mAs, 120 kV). Radiation dose was measured by assessing the volumetric CT dose index (CTDI_vol), dose length product (DLP) and effective dose (ED). Signal-to-noise ratio (SNR) and image noise were assessed for objective image quality, visibility of arteries and grey-white differentiation for subjective image quality.

RESULTS

Two hundred patients (n = 100 in each group) were included. In group A, median selected tube current was 643 mAs (group B, 300 mAs; p < 0.001). Median values of CTDI_vol, DLP and ED were 91.54 mGy, 1561 mGy cm and 2.97 mSv in group A, and 43.15 mGy, 769 mGy cm and 1.46 mSv in group B (p < 0.001). Image quality did not significantly differ between groups (p > 0.05).

CONCLUSIONS

ATCM yielded disproportionally high radiation dose due to substantial tube current increase at the frame level, while image quality did not improve. Thus, ATCM should preferentially be disabled.

KEY POINTS

• Automated tube current modulation (ATCM) yields disproportionally high radiation dose in intraoperative CT angiography of the brain with stereotactic head frames. • ATCM does not improve overall image quality in these special examinations. • ATCM is not yet optimised for CT angiography of the brain with major extracorporeal foreign materials within the scan range.

Automatic tube potential selection with tube current modulation in coronary CT angiography: Can it achieve consistent image quality among various individuals?

The present study included a total of 111 consecutive patients who had undergone coronary computed tomography (CT) angiography, using a first-generation dual-source CT with automatic tube potential selection and tube current modulation. Body weight (BW) and body mass index (BMI) were recorded prior to CT examinations. Image noise and attenuation of the proximal ascending aorta (AA) and descending aorta (DA) at the middle level of the left ventricle were measured. Correlations between BW, BMI and objective image quality were evaluated using linear regression. In addition, two subgroups based on BMI (BMI ≤25 and >25 kg/m²) were analyzed. Subjective image quality, image noise, the signal-to-noise ratio (SNR) and the contrast-to-noise ratio (CNR) were all compared between those. The image noise of the AA increased with the BW and BMI (BW: r=0.453, P<0.001; BMI: r=0.545, P<0.001). The CNR and SNR of the AA were inversely correlated with BW and BMI, respectively. The image noise of the DA and the CNR and SNR of the DA exhibited a similar association to those with the BW or BMI. The BMI >25 kg/m² group had a significant increase in image noise (33.1±6.9 vs. 27.8±4.0 HU, P<0.05) and a significant reduction in CNR and SNR, when compared with those in the BMI ≤25 kg/m² group (CNR: 18.9±4.3 vs. 16.1±3.7, P<0.05; SNR: 16.0±3.8 vs. 13.6±3.2, P<0.05). Patients with a BMI of ≤25 kg/m² had more coronary artery segments scored as excellent, compared with patients with a BMI of >25 kg/m² (P=0.02). In conclusion, this method is not able to achieve a consistent objective image quality across the entire patient population. The impact of BW and BMI on objective image quality was not completely eliminated. BMI-based adjustment of the tube potential may achieve a more consistent image quality compared to automatic tube potential selection, particularly in patients with a larger body habitus.