RPS 1001a - Advances in MRI techniques

Purpose:

To evaluate intrapatient and interobserver variability in patients who underwent liver MRI with Gd-EOB-DTPA using two different multiarterial phase techniques.

Methods and materials:

Two radiologists retrospectively analysed 154 prospectively enrolled patients who underwent liver MRI performed with 1.5T MAGNETOM Aera twice from February 2017 to March 2019, using two different multiarterial algorithms: CAIPIRINHA or TWIST-VIBE (Siemens Healthcare, Erlangen, Germany). For every patient, breath-holding time, BMI, sex, and age were recorded before the second MRI study.

The phase without contrast media and every arterial phase (after administration of Gd-EOB-DTPA; dose: 0.025ml/kg; injection rate: 1ml/s) were evaluated using different scores to quantify Gibbs artefacts (present/absent), noise (present/absent), breath artefacts, and general quality of images (score 1-5, 1: “non-diagnostic”, 5: “absence of artefacts”/“optimal exam quality”).

Results:

CAIPIRINHA always had better scores for every parameter considered, with the exception of noise score analysis. 37% of patients failed to obtain multiarterial phases: in particular 11% had only one arterial phase, 26% had two. Breath-holding time was the only parameter that influenced the performance of multiarterial techniques.

TWIST-VIBE had a worse score for Gibbs and breath artefacts but a better noise score: 23% of patients in the basal phase had noise (vs46% in CAIPIRINHA), 18% in the first arterial phase (vs29%), 18% in the second arterial phase (vs30%), and 17% in the third arterial phase (vs31%).

Conclusion:

CAIPIRINHA always allows the obtaining of diagnostic images without multiarterial phase only in 37% of cases. TWIST-VIBE always allows the obtaining of three arterial phases.

Breath-holding time is the only parameter which can influence the preliminary choice between CAIPIRINHA and TWIST-VIBE.

TWIST-VIBE is preferable in very compliant patients, otherwise CAIPIRINHA is more appropriate.

Limitations:

Limited cohort; non-homogeneous follow-up.

Ethics committee approval

Study approved by ethics committee.

Funding:

No funding was received for this work.

Purpose:

To determine whether T1 reduction rate is dependent on field strength in patients with and without liver cirrhosis.

Methods and materials:

85 consecutive Gd-EOB-DTPA liver MRI scans with available T1 mapping pre- and post-contrast administration in equilibrium phase were analysed between September 2018 and January 2019. 44 exams were performed on a 1.5T system and 41 exams on a 3T system. Each of these two groups was then divided into patients with and without liver cirrhosis. T1-reduction rates were calculated as: (native T1 – post contrast T1) / native T1. Groups were compared using a Mann-Whitney-U test.

Results:

At 1.5, 23 patients had cirrhosis, while 21 patients had no cirrhosis. At 3T, 17 patients had cirrhosis, while 24 patients had no cirrhosis. At both 3T and 1.5T, T1 reduction rate discriminated between patients with and without cirrhosis (p=0.001 and p>0.001). T1 reduction rates did not differ significantly between 3T and 1.5T in patients with cirrhosis (median 60% vs. 59%; p=0.991) and without cirrhosis (median 75% vs. 75%; p=0.787).

Conclusion:

This study shows that T1 reduction rate is comparable between 3T and 1.5T and may discriminate patients with and without liver cirrhosis as an imaging surrogate to map liver function.

Limitations:

We acknowledge limitations to our study, mainly the retrospective study design and limited number of patients included.

Ethics committee approval

This study was approved by the cantonal ethics committee Bern. Written informed consent was given.

Funding:

This project was funded by the Swiss National Science Foundation (SNF) grant # 320030_188591.

Purpose:

A quantitative MRI method (QMRI) was recently validated using multi-compartmental R2 relaxometry (MCR2R) to map liver fibrosis/inflammation in patients with non-alcoholic steatohepatitis. The study was designed to evaluate this method precision through examinations performed in volunteers on a 3T scanner at different times using the same parameters.

Methods and materials:

30 volunteers were recruited. They underwent exams on a 3T scanner after fasting for 4h. They were re-scanned on the same MRI and protocol less than 1 week apart. The protocols used a multi-spin echo sequence (MSE) to determine water extracellular fraction and R2 maps. We also acquired multi-gradient echo sequences (MGE) to calculate proton density fat fraction. All measures were performed by one radiologist with 13 years of experience using a ROI encompassing the whole liver area in a single slice. The repeatability of the measures was assessed through the within-subject coefficient of variation and repeatability coefficient.

Results:

Extracellular water fraction (ECWF) mean values were 12.99% (±1.88) in the first exam and 13.06 (±1.9) in the second, ranging from 9.9 to 16.9%. The repeated scans were not statistically significantly different (T-test). The bias between scans was <0.1% ECWF. The repeatability coefficient between scans was 2.47% (absolute measure of ECWF), comparable to the 2xSD variation of ECWF obtained previously across each fibrosis grade, indicating that approximately 95% of patients would receive the correct fibrosis grade. The within-subject coefficient of variation was 6.86% (as a percentual).

Conclusion:

The study showed good precision of the QMRI method for quantifying the extracellular water fraction (as a proxy for degree of fibrosis), with repeatability coefficient of 2.47%.

Limitations:

Single institution. Small number of patients. One reader.

Ethics committee approval

Approved by ethics with informed consent.

Funding:

No funding was received for this work.

Purpose:

To investigate the diagnostic value of different DWI models (monoexponential, biexponential models multiple b values diffusion-weighted imaging, and non-Gaussian diffusion-weighted MRI) in poorly differentiated pancreatic ductal adenocarcinoma.

Methods and materials:

52 patients with poorly differentiated pancreatic ductal adenocarcinoma confirmed by surgery were collected. All patients underwent DWI (1.5T, Multi-b values：0, 50, 100, 150, 200, 500, 800, 1000, 1500, 2000 s/mm2). Mean values of DWI-derived metrics ADC_standard, ADC_slow, ADC_fast, f, MK, and DK were calculated from regions of interest in all tumours and non-tumorous parenchyma and compared. ROC was used to evaluate the diagnostic efficiency.

Results:

All lesions could be distinguished from non-lesions by three DWI sequences (ADC-DWI, DKI and IVIM). The ADC_standard, ADC_fast , f values, and MK showed significant differences between tumours and non-tumorous parenchyma (both P < 0.05). The area under the curve for ADC, D, D*, f, K, and DK were (0.705, 0.665, 0.648, 0.614), respectively. The ROC curve integrated with ADC_standard and MD has better diagnostic efficiency (the area under ROC curve is about 0.754).

Conclusion:

ADC_standard, ADC_fast, f, and MK could differentiate tumours from non-tumorous parenchyma. The combination of Gaussion distribution model and non-Gaussion distribution model has the potential to increase the diagnostic accuracy of DWI in patients with pancreatic ductal adenocarcinoma.

Limitations:

Not enough cases and focused only on poorly differentiated pancreatic cancer.

Ethics committee approval

n/a

Funding:

Sichuan science and technology program（2015SZ0030. Popularisation project of sichuan provincial commission of health and family planning（17PJ421.

Purpose:

To determine the accuracy of the high specificity of hyperintensity foci presence on T1-weighted fat-suppressed images in the diagnosis of superficial and deep pelvic endometriosis in everyday practice.

Methods and materials:

During 12 months, 40 patients (35.4 ± 7.2 years) referred to our institution with clinically suspected endometriosis underwent 3.0T MRI examination of pelvis without using i.v. contrast. Every examination included T1-weighted fat-sat sequences and was interpreted by two radiologists (with different experience). All patients with T1-W fat-sat hyperintensity foci proceded afterwards to minimally invasive diagnostic procedures as laparoscopy or cystoscopy for confirmation of endometriosis with histopathology.

Results:

The two radiologists had an identical sensitivity of 80.3% (95% CI; 57.6-95.2). The specificity was 100% (95% CI; 82.4-100) for the young and 89.5% (95% CI; 66.9-98.7) for the experienced radiologist. The area under the ROC curve was 0.90 (95% CI; 0.81-0.99) for the young and 0.83 (95% CI; 0.70-0.96) for the experienced radiologist. The intraoperator variability was low with almost perfect reproducibility for the overall detection of lesions; k=0.90 (95% CI; 0.77-1) for the young and k=0.85 (95% CI; 0.70-1) for the experienced radiologist.

Conclusion:

Our results show that the characterisation of lesions by MRI, especially with using of T1-W fat sat sequence, is highly specific and moderately sensitive in the diagnosis of endometriosis, with high interoperator reproducibility. Pelvic protocols should always include the T1-W fat-sat sequence. In addition to its accuracy in distinguishing teratomas from endometriomas, it helps to better characterise adhesions and distinguishing from post-surgical scars. In future MRIs, it can be used to diagnose pelvic lesions and the accurate examination may replace the necessity of biopsy.

Limitations:

n/a

Ethics committee approval

n/a

Funding:

No funding was received for this work.

Purpose:

This study compared enhancement of the portal vein with Gd-BOPTA and Gd-EOB-DTPA in cirrhotic patients.

Methods and materials:

A total of 84 cirrhotic patients (30 women, 54 men) who underwent both Gd-EOB-DTPA– and Gd-BOPTA–enhanced liver MRI from 2008 to 2018 were retrospectively included. Interval time between Gd-BOPTA and Gd-EOB-DTPA MR studies was 10.7 months. Patients received a weight-dependent dose of Gd-EOB-DTPA (0.025 mmol/kg body) or Gd-BOPTA (0.1 mmol/kg body). Signal intensity of the portal vein, liver-to-portal vein contrast ratio, and image contrast of portal vein were calculated on portal venous phase for each exam and compared per patient through the Wilcoxon signed rank sum test. Statistical significance was set at p<0.05.

Results:

Compared to Gd-BOPTA MRI, Gd-EOB MRI in the portal venous phase showed significantly lower portal vein signal intensity (1461.72 ± 360.56 vs. 1256.68 ± 327.71, respectively; p<0.001) and image contrast of the portal vein (0.27 ± 0.11 vs. 0.35 ± 0.11, respectively, p<0.001) and higher liver-to-portal vein contrast ratio (0.72 ± 0.11 vs. 0.64 ± 0.11, p<0.001).

Conclusion:

At the recommended dose of hepatobiliary contrast agents, Gd-EOB-DTPA yields lower enhancement of the portal vein than Gd-BOPTA in cirrhotic patients. Gd-BOPTA might therefore enable better evaluation of the portal vein in cirrhotic patients.

Limitations:

Retrospective design, small sample size and single-centre experience, lack of correlation with laboratory data or contrast dose.

Ethics committee approval

IRB approved, informed consent waived.

Funding:

No funding was received for this work.