Research Presentation Session

RPS 1101 - Advances in CT techniques

Lectures

1
RPS 1101 - Comparison study between published contrast administration protocols for enhanced liver CT examination in adults

RPS 1101 - Comparison study between published contrast administration protocols for enhanced liver CT examination in adults

08:56F. Zanca, Leuven / BE

Purpose:

To compare hepatic enhancement level and uniformity across adult patients, using different personalised contrast administration protocols proposed in literature.

Methods and materials:

623 patients undergoing a three-phase liver CT examination were prospectively included, using a standardised CT protocol (100 or 120 kV in the function of patient BMI, 5 ml/sec contrast media (CM) injection rate, 80 sec portal phase delay, Iopamidol 370mg Iodine/ml).

Patients were prospectively assigned to one of three specific CM administration protocol group: G1 (100ml fixed CM volume, n = 297); G2 (600mgI/kg of body weight, n = 98,); G3 (750mgI/kg of measured fat free mass (FFM), n=228). Impact of kV settings was accounted through measured liver parenchymal enhancement (CEI) scaling factors for each scanner.

The mean injected CM volume, iodine dose, and median CEI at 100kV were compared across groups (one-way ANOVA or Kruskall-Wallis, p<0.005). The variance of the distributions was assessed and CEI compared to a diagnostically appropriate level of 30-50HU.

Results:

G3 mean contrast volume and iodine dose (93.1 ml, 34.4 mgI respectively) were significantly lower than G1 (100ml, 37.0 mgI) and G2 (109.5 ml, 40.5 mgI) (p<0.0001 for both). 87-94% of patients were over-enhanced with a significantly lower median CEI for G3 (65.6HU) than G2 (70.7HU) and G1 (77.5HU) (p<0.0001 for both). Only 8% (G1), 6% (G2), and 13% (G3) reached target enhancement. Variance decreased from 520 (G1) to 403 (G2) to 247 (G3).

Conclusion:

The FFM-based protocol improved patient-to-patient liver enhancement uniformity, while significantly reducing the iodine load, although still overestimated for a large group of patients. Our kV scaling factors suggest that 600mgI/FFM@100kV would significantly reduce over-enhancement and target a homogeneous diagnostic value of 50 HU.

Limitations:

n/a

Ethics committee approval

Approved.

Funding:

No funding was received for this work.

2
RPS 1101 - Contrast media reduction in abdominal dual-energy CT: low keV virtual monoenergetic images restore diagnostic assessment and image quality

RPS 1101 - Contrast media reduction in abdominal dual-energy CT: low keV virtual monoenergetic images restore diagnostic assessment and image quality

06:29S. Lennartz, Köln / DE

Purpose:

Abdominal CT with reduced contrast media (CM) application would be beneficial for patients at risk for contrast-induced nephropathy yet may imply inferior assessability. The study evaluated if low-keV virtual monoenergetic images from abdominal spectral-detector CT (SDCT) with reduced contrast media (RCM-VMI40keV) provide similar image quality as conventional scans with standard contrast media dose (SCM).

Methods and materials:

78 patients with abdominal SDCT were included: 41 patients at risk for adverse reactions who received 44 RCM scans with 50ml of CM and 37 patients who received 44 SCM scans with 100ml of CM. Both groups were matched for effective body diameter. RCM-VMI40keV, conventional RCM, and SCM images were reconstructed. Attenuation and SNR of liver, pancreas, kidneys, lymph nodes, aorta, and portal vein were assessed ROI-based. CNR of lymph nodes vs aorta/portal vein were calculated. Two readers blinded against patients/reconstructions assessed organ/vessel contrast, lymph node delineation, image noise, and overall assessability using 4-point Likert scales.

Results:

Lymph node attenuation was similar between RCM-VMI40keV andSCM images (p=0.83), while for all other ROIs, RCM-VMI40keV was superior (p<0.05). SNR was equal between RCM-VMI40keV and SCM images for all ROIs (p-range 0.23-0.99). CNR of lymph nodes vs aorta/portal vein was highest in RCM-VMI40keV (p<0.05). Qualitatively, RCM-VMI40keV received equivalent or higher scores than SCM in all criteria except for organ contrast, overall assessability, and subjective image noise for which SCM was superior. However, in these three categories, RCM-VMI40keV received proper or excellent scores in 88.6%/94.2%/95.4% of all cases. Conventional RCM were inferior in all quantitative/qualitative parameters.

Conclusion:

VMI40keV effectively antagonises contrast deterioration in CM-reduced abdominal SDCT, facilitating diagnostic assessment.

Limitations:

The study did not investigate lesion detection.

Ethics committee approval

Waived due to retrospective study characteristics.

Funding:

Else Kröner-Fresenius Stiftung (2016-Kolleg-19 to SL).

3
RPS 1101 - Optimised virtual monoenergetic image for liver fibrosis staging using dual-layer spectral CT

RPS 1101 - Optimised virtual monoenergetic image for liver fibrosis staging using dual-layer spectral CT

05:33R. Li, Shanghai / CN

Purpose:

To investigate the value of an optimised virtual monoenergetic image (VMI) for liver fibrosis staging using dual-layer spectral CT.

Methods and materials:

26 rabbit models of CCl4-induced liver fibrosis were established and 4 untreated rabbits served as controls. Dynamic contrast-enhanced CT was performed including arterial phase [AP] and venous phase [VP] using dual-layer spectral CT. CT attenuation on VMI (40kev, 50kev, 60kev, 70kev) and conventional polyenergetic images (CPI) were measured by whole-liver volumetric ROI drawing on precontrast and enhancement images. ΔCT40-70kev were calculated and correlated with the histopathological fibrosis stage to determine the optimised VMI.

Receiver operating characteristic (ROC) analysis was performed for assessing the diagnostic performance of the optimised VMI for fibrosis stage.

Results:

On arterial phase, no significant correlation was identified between ΔCT40-70kev and fibrosis stage or CPA (P>0.05). On portal venous phase, ΔCTCPI, ΔCTVMI 40, ΔCTVMI 50, ΔCTVMI 60, and ΔCTVMI 70 showed moderate correlation (r=0.453, P=0.012; r=0.595, P=0.001; r=0.543, P=0.002; r=0.498, P=0.005; and r=0.449, P=0.013; respectively) with fibrosis. ΔCTVMI 50 showed the highest correlation with fibrosis stage. ΔCTVMI 50 areas under ROC (AUROCs) were 0.861, 0.790, 0.831, and 0.740 for diagnosing fibrosis with ≥F1, ≥F2, ≥F3, and F4 stage, respectively.

Conclusion:

ΔCTVMI 50 on portal venous phase may be a potential biomarker for liver fibrosis staging.

Limitations:

This is an anmial study. The study should be further confirmed in patients.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

4
RPS 1101 - Metal implants on abdominal CT: can split-filter dual-energy CT provide additional value over iterative metal artefact reduction?

RPS 1101 - Metal implants on abdominal CT: can split-filter dual-energy CT provide additional value over iterative metal artefact reduction?

05:55H. Wichtmann, Basel / CH

Purpose:

To assess metal artefact reduction in split-filter dual-energy CT (tbDECT) using virtual monoenergetic images (VMI) compared to 120kVp-equivalent mixed images (MIX) and iterative metal artefact reduction algorithm (iMAR).

Methods and materials:

Abdominal tbDECT of 30 patients with total hip replacements (15 uni-, 15 bilateral) were included. Images were reconstructed as MIX and VMI (40 to 190 keV, 10 keV increments), with and without iMAR. Quantitative image quality was assessed using ROI-analysis of corrected attenuation-values [HU] and standard deviation (SD) for hypo- and hyperdense artefact on all reconstructions. Qualitative image quality was rated for overall image quality and vascular contrast on a 5-point Likert-scale (1: no artefact/contrast to 5: most artefact/contrast).

Results:

Lowest artefact (both hypo- and hyperdense) was observed on MIXiMAR (-10.6 and -0.6 HU), which was significantly lower compared to the standard MIX (p=0.006 and p<0.001). Comparison between MIX and VMI40keV showed more artefact on VMI40keV (p<0.001). Low keV VMIiMAR-40keV did not show more artefact than MIXiMAR or MIX (p≥0.08). Image noise was highest for conventional low VMI40keV (46.1 HU). Yet, VMIiMAR-40keV showed similar image noise compared to MIX (10.9 vs 12.4 HU, p=0.608). Qualitative image quality was rated best for VMIiMAR at 140 keV (2.43), which was significantly better than MIX (4.07, p<0.001), but not MIXiMAR (2.70, p=0.176). Low keV VMIiMAR showed both improved image quality and vascular contrast compared to MIX (2.93 vs 4.07, p<0.001, 4.17 vs 2.70, p<0.001).

Conclusion:

Abdominal tbDECT MIXiMAR reconstructions provide the best image quality in patients with total hip replacements. High keV VMIiMAR does not further decrease artefact. However, iMAR enables the use of low keV VMIiMAR to boost iodine-based contrast with acceptable image quality.

Limitations:

n/a

Ethics committee approval

IRB approved, need for informed consent waived.

Funding:

No funding was received for this work.

5
RPS 1101 - A randomised controlled trial proposing a straight forward 10-to-10 rule for individualised liver imaging based on tube voltage and body weight

RPS 1101 - A randomised controlled trial proposing a straight forward 10-to-10 rule for individualised liver imaging based on tube voltage and body weight

07:02B. Martens, Maastricht / NL

Purpose:

To optimise the CT liver protocol based on both body weight (BW) and tube voltage (TV). The rule of thumb was that 10kV TV reduction would lead to a 10% decrease in CM dose.

Methods and materials:

256 patients referred for a portal-venous-phase CT were randomly allocated to one of four groups. Group 1 (n=64): 120kV; 0.521gI/kg. In group 2 (n=63), TV was reduced (90kV), whereas the dosing factor remained unaltered: 0.521gI/kg. In group 3 (n=63), TV was reduced by 20kV with a subsequent 20% reduction in dosing factor (e.g. 100kV; 0.415gI/kg). Group 4 (n=66): 30 kV decrease with a 30% CM reduction: 90kV; 0.365gI/kg. Objective image quality (IQ) was evaluated by measuring the attenuation in Hounsfield units (HU), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR). Subjective IQ was assessed by using a 5-point Likert scale regarding overall IQ. Statistical analysis was performed using SPSS (IBM, version 24.0).

Results:

Mean attenuation values in group 1, 3, and 4 were comparable: 118.2±10.0; 117.6±13.9; 117.3±21.6 (p>0.90). Attenuation in group 2 was significantly higher, 141.0±18.2, in comparison to all other groups (p=<0.01). Patients were divided into two weight categories: 80kg and >80kg. No significant difference in attenuation was found between weight categories. CNR was significantly higher in group 2 compared to the other three groups (p=<0.01). No significant differences in subjective IQ were found (p=0.383).

Conclusion:

The proposed 10-to-10 rule is an easy to reproduce method leading to homogeneous enhancement of the liver, regardless of BW and TV.

Limitations:

This is a single-centre study.

Ethics committee approval

Approved by the local ethics committee. Registered on ClinicalTrials.gov (NCT03735706). Written informed consent was obtained from all 256 patients.

Funding:

No funding was received for this work.

6
RPS 1101 - Dynamic segmental CT liver perfusion data analysis after portosystemic shunt procedure in patients with liver cirrhosis

RPS 1101 - Dynamic segmental CT liver perfusion data analysis after portosystemic shunt procedure in patients with liver cirrhosis

06:12N. Djuraeva, Tashkent / UZ

Purpose:

To study the dynamics of parameters of the CT liver perfusion in patients with liver cirrhosis after portosystemic shunting.

Methods and materials:

Study included 50 patients (average age 43±1.5 years). The control group consisted of 10 healthy volunteers of comparable age. All patients underwent volumetric low-dose liver perfusion with a tube rotation speed of 0.275 sec, the amount of contrast medium 40 ml, Kv 100/80, MA 200/150, and effective dose (E) 17±1,2 mSv.

Results:

In the group with liver cirrhosis (LC), there was an increase in global and regional changes in hepatic perfusion towards an increase in hepatic arterial blood flow (BF) and a decrease in portal BF with a decrease in the hepatic perfusion index (HPI). After portosystemic shunting, there was a positive shift in the portal BF by 87,44±6,1 (mL/100 mL/min) and HPI improvement by 31,4 % (p<0.05). There was also a decrease in the width of the portal and splenic veins by 17.1% and 16.5% (p<0.05), respectively.

Segmental analysis of HPI showed that the most pronounced improvement in hepatic blood flow was observed in I, IV, VII, VIII segments (17.4%).

Conclusion:

Low-dose CT liver perfusion revealed significant positive shifts (p<0.05) after portosystemic shunting of the HPI index, more pronounced in central segments in patients with LC, restoring the balance between hepatic and portal blood flow, and reducing the pressure in the portal and splenic veins.

Limitations:

n/a

Ethics committee approval

n/a

Funding:

No funding was received for this work.

PEP Subscription Required

This course is only accessible for ESR Premium Education Package subscribers.