Predictive Value of Cardiac CTA, Cardiac MRI, and Transthoracic Echocardiography for Cardioembolic Stroke Recurrence.

BACKGROUND. Transthoracic echocardiography (TTE) is the standard of care for initial evaluation of patients with suspected cardioembolic stroke. Although TTE is useful for assessing certain sources of cardiac emboli, its diagnostic capability is limited in the detection of other sources, including left atrial thrombus and aortic plaques. OBJECTIVE. The purpose of this article was to investigate sensitivity, specificity, and predictive value of cardiac CTA (CCTA), cardiac MRI (CMRI), and TTE for recurrence in patients with suspected cardioembolic stroke. METHODS. We retrospectively included 151 patients with suspected cardioembolic stroke who underwent TTE and either CMRI (n = 75) or CCTA (n = 76) between January 2013 and May 2017. We evaluated for the presence of left atrial thrombus, left ventricular thrombus, vulnerable aortic plaque, cardiac tumors, and valvular vegetation as causes of cardioembolic stroke. The end point was stroke recurrence. Sensitivity, specificity, PPV, and NPV for recurrent stroke were calculated; the diagnostic accuracy of CMRI, CCTA, and TTE was compared between and within groups using AUC. RESULTS. Twelve and 14 recurrent strokes occurred in the CCTA and CMRI groups, respectively. Sensitivity, specificity, PPV, and NPV were 33.3%, 93.7%, 50.0%, and 88.2% for CCTA; 14.3%, 80.3%, 14.3%, and 80.3% for CMRI; 14.3%, 83.6%, 16.7%, and 80.9% for TTE in the CMRI group; and 8.3%, 93.7%, 20.0%, and 84.5% for TTE in the CCTA group. Accuracy was not different (p > .05) between CCTA (AUC = 0.63; 95% CI, 0.49-0.77), CMRI (0.53; 95% CI, 0.42-0.63), TTE in the CMRI group (0.51; 95% CI, 0.40-0.61), and TTE in the CCTA group (0.51; 95% CI, 0.42-0.59). In the CCTA group, atrial and ventricular thrombus were detected by CCTA in three patients and TTE in one patient; in the CMRI group, thrombus was detected by CMRI in one patient and TTE in two patients. CONCLUSION. CCTA, CMRI, and TTE showed comparably high specificity and NPV for cardioembolic stroke recurrence. CCTA and CMRI may be valid alternatives to TTE. CCTA may be preferred given potentially better detection of atrial and ventricular thrombus. CLINICAL IMPACT. CCTA and CMRI have similar clinical performance as TTE for predicting cardioembolic stroke recurrence. This observation may be especially important when TTE provides equivocal findings.

Agreement between magnetic resonance imaging and computed tomography in the postnatal evaluation of congenital lung malformations: a pilot study.

OBJECTIVES: To compare postnatal magnetic resonance imaging (MRI) with the reference standard computed tomography (CT) in the identification of the key features for diagnosing different types of congenital lung malformation (CLM). METHODS: Respiratory-triggered T2-weighted single-shot turbo spin echo (ss-TSE), respiratory-triggered T1-weighted turbo field echo (TFE), balanced fast field echo (BFFE), and T2-weighted MultiVane sequences were performed at 1.5 T on 20 patients prospectively enrolled. Two independent radiologists examined the postnatal CT and MRI evaluating the presence of cysts, hyperinflation, solid component, abnormal arteries and/or venous drainage, and bronchocele. Diagnostic performance of MRI was calculated and the agreement between the findings was assessed using the McNemar-Bowker test. Interobserver agreement was measured with the kappa coefficient. RESULTS: CT reported five congenital pulmonary airway malformations (CPAMs), eight segmental bronchial atresias, five bronchopulmonary sequestrations (BPS), one congenital lobar overinflation, one bronchogenic cyst, and three hybrid lesions. MRI reported the correct diagnosis in 19/20 (95%) patients and the malformation was correctly classified in 22/23 cases (96%). MRI correctly identified all the key findings described on the CT except for the abnormal vascularization (85.7% sensitivity, 100% specificity, 100% PPV, 94.1% NPV, 95% accuracy for arterial vessels; 57.1% sensitivity, 100% specificity, 100% PPV, 84.2% NPV, 87% accuracy for venous drainage). CONCLUSIONS: MRI can represent an effective alternative to CT in the postnatal assessment of CLM. In order to further narrow the gap with CT, the use of contrast material and improvements in sequence design are needed to obtain detailed information on vascularization, which is essential for surgical planning. KEY POINTS: • Congenital lung malformations (CLMs) can be effectively studied by MRI avoiding radiation exposure. • Crucial features of CLM have similar appearance when comparing CT with MRI. • MRI performs very well in CLM except for aberrant vessel detection and characterization.

High-quality low-dose cardiovascular computed tomography (CCT) in pediatric patients using a 64-slice scanner.

Background Cardiovascular computed tomography (CCT) technology is rapidly advancing allowing to perform good quality examinations with a radiation dose as low as 1.2 mSv. However, latest generation scanners are not available in all centers. Purpose To estimate radiation dose and image quality in pediatric CCT using a standard 64-slice scanner. Material and Methods A total of 100 patients aged 6.9 ± 5.4 years (mean ± standard deviation) who underwent a 64-slice CCT scan using 80, 100, or 120 kVp, were retrospectively evaluated. Radiation effective dose was calculated on the basis of the dose length product. Two independent readers assessed the image quality through signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and a qualitative score (3 = very good, 2 = good, 1 = poor). Non-parametric tests were used. Results Fifty-five exams were not electrocardiographically (ECG) triggered, 20 had a prospective ECG triggering, and 25 had retrospective ECG triggering. The median effective dose was 1.3 mSv (interquartile range [IQR] = 0.8-2.7 mSv). Median SNR was 30.6 (IQR = 23.4-33.6) at 120 kVp, 29.4 (IQR = 23.7-34.8) at 100 kVp, and 24.7 (IQR = 19.4-34.3) at 80 kVp. Median CNR was 21.0 (IQR = 14.8-24.4), 19.1 (IQR = 15.6-23.9), and 25.3 (IQR = 19.4-33.4), respectively. Image quality was very good, good, and poor in 56, 39, and 5 patients, respectively. No significant differences were found among voltage groups for SNR ( P = 0.486), CNR ( P = 0.336), and subjective image quality ( P = 0.296). The inter-observer reproducibility was almost perfect (κ = 0.880). Conclusion High-quality pediatric CCT can be performed using a 64-slice scanner, with a radiation effective dose close to 2 mSv in about 50% of the cases.

Quantitative analysis of dynamic computed tomography angiography for the detection of endoleaks after abdominal aorta aneurysm endovascular repair: A feasibility study.

OBJECTIVES: To assess the feasibility of quantitative analysis of dynamic computed tomography angiography (dCTA) for the detection of endoleaks in patients who underwent endovascular repair of abdominal aortic aneurysms (EVAR). MATERIAL AND METHODS: Twenty patients scheduled for contrast-enhanced CT angiography (CTA) of the abdominal aorta post-EVAR were prospectively enrolled. All patients received a standard triphasic CTA protocol, followed by an additional dCTA. The dCTA acquisition enabled reconstruction of color-coded maps depicting blood perfusion and a dCTA dataset of the aneurysm sac. Observers assessed the dCTA and dynamic CT perfusion (dCTP) images for the detection of endoleaks, establishing diagnostic confidence based on a modified 5-point Likert scale. An index was calculated for the ratio between the endoleak and aneurysm sac using blood flow for dCTP and Hounsfield units (HU) for dCTA. The Wilcoxon test compared the endoleak index and the diagnostic confidence of the observers. RESULTS: In total, 19 patients (18 males, median age 74 years [70.5-75.7]) were included for analysis. Nine endoleaks were detected in 7 patients using triphasic CTA as the reference standard. There was complete agreement for endoleak detection between the two techniques on a per-patient basis. Both dCTA and dCTP identified an additional endoleak in one patient. The diagnostic confidence using dCTP for detection of endoleaks was not significantly superior to dCTA (5.0 [5-5] vs. 4.5 [4-5], respectively; p = 0.11); however, dCTP demonstrated superior diagnostic confidence for endoleak exclusion compared to dCTA (1.0 [1-1] vs 1.5 [1.5-1.5], respectively; p <0.01). Moreover, the dCTP endoleak index was significantly higher than the dCTA index (18.5 [10.8-20.5] vs. 3.5 [5-2.7], respectively; p = 0.02). CONCLUSIONS: Quantitative analysis of dCTP imaging can aid in the detection of endoleaks and demonstrates a higher endoleak detection rate than triphasic CTA, as well as a strong correlation with visual assessment of dCTA images.

Fully automated contour detection of the ascending aorta in cardiac 2D phase-contrast MRI.

PURPOSE: In this study we proposed a fully automated method for localizing and segmenting the ascending aortic lumen with phase-contrast magnetic resonance imaging (PC-MRI). MATERIAL AND METHODS: Twenty-five phase-contrast series were randomly selected out of a large population dataset of patients whose cardiac MRI examination, performed from September 2008 to October 2013, was unremarkable. The local Ethical Committee approved this retrospective study. The ascending aorta was automatically identified on each phase of the cardiac cycle using a priori knowledge of aortic geometry. The frame that maximized the area, eccentricity, and solidity parameters was chosen for unsupervised initialization. Aortic segmentation was performed on each frame using active contouring without edges techniques. The entire algorithm was developed using Matlab R2016b. To validate the proposed method, the manual segmentation performed by a highly experienced operator was used. Dice similarity coefficient, Bland-Altman analysis, and Pearson's correlation coefficient were used as performance metrics. RESULTS: Comparing automated and manual segmentation of the aortic lumen on 714 images, Bland-Altman analysis showed a bias of -6.68mm2, a coefficient of repeatability of 91.22mm2, a mean area measurement of 581.40mm2, and a reproducibility of 85%. Automated and manual segmentation were highly correlated (R=0.98). The Dice similarity coefficient versus the manual reference standard was 94.6±2.1% (mean±standard deviation). CONCLUSION: A fully automated and robust method for identification and segmentation of ascending aorta on PC-MRI was developed. Its application on patients with a variety of pathologic conditions is advisable.

Strain of ascending aorta on cardiac magnetic resonance in 1027 patients: Relation with age, gender, and cardiovascular disease.

OBJECTIVES: To evaluate ascending aortic strain (AAS) with cardiac magnetic resonance (CMR) in a large consecutive series of patients with different types of cardiovascular disease (CVD). METHODS: Two-dimensional phase-contrast gradient-echo sequences of the ascending aorta were retrospectively reviewed in 1027 patients (726 males, 301 females). Aortic lumen area was segmented using a semi-automatic approach to calculate AAS values. Subgroup analysis was performed for patients with normal CMR, tetralogy of Fallot (ToF), and ischemic heart disease (IHD). Multivariate and post-hoc analyses were performed to evaluate the effect of age, gender, and CVD on AAS values. Shapiro-Wilk, three- and two-way ANOVA, Mann-Whitney U, and Spearman correlation statistics were used. RESULTS: Multivariate analysis showed significant differences in AAS among decades of age (p<0.001), genders (p=0.006) and CVD subgroups (p<0.001) without interaction among these factors. A gender-related difference (higher AAS in females) was significant in ToF (p=0.008), while an AAS reduction during aging was observed in all CVD subgroups. Post-hoc analysis showed a significantly lower AAS in ToF and IHD patients compared to subjects with normal CMR (p<0.001). CONCLUSION: Differences in age, gender, and CVD independently affect AAS. The lower AAS observed in ToF fosters its assessment during follow-up in adulthood. Future studies on causes and clinical implications of a higher AAS in females affected by ToF are warranted.