Auxiliary diagnostic potential of ventricle geometry and late gadolinium enhancement in left ventricular non-compaction; non-randomized case control study

Background

There are still ambiguities existing in regard to left ventricular non-compaction (LVNC) diagnostic imaging. The aim of our study was to analyze diagnostic potential of late gadolinium enhancement (LGE) and ventricle geometry in patients with LVNC and controls.

Methods

Data on cardiac magnetic resonance imaging (CMR) studies for LVNC were reassessed from the hospital’s database (3.75 years; n=1975 exams). Matching sample of controls included cases with no structural heart disease, hypertrophic or dilative cardiomyopathy, arrhythmogenic right ventricular dysplasia or subacute myocarditis. Eccentricity of the left ventricle was measured at end diastole in the region with pronounced NC and maximal to minimal ratio (MaxMinEDDR) was calculated.

Results

Study included 255 patients referred for CMR, 100 (39.2%) with LVNC (prevalence in the studied period 5.01%) and 155 (60.8%) controls. Existing LGE had sensitivity of 52.5% (95%-CI:42.3–62.5), specificity of 80.4% (95%-CI:73.2–86.5) for LVNC, area under curve (AUC) 0.664 (95%-CI:0.603–0.722);p<0.001. MaxMinEDDR>1.10 had sensitivity of 95.0% (95%-CI:88.7–98.4), specificity of 82.6% (95%-CI: 75.7–88.2) for LVNC, AUC 0.917 (95%-CI:0.876–0.948); p<0.001. LGE correlated with Max-Min-EDD-R (Rho=0.130; p=0.038) and there was significant difference in ROC analysis ΔAUC0.244 (95%-CI:0.175–0.314); p<0.001. LGE also correlated negatively with stroke volume and systolic function (both p<0.05, respectively).

Conclusions

LGE was found to be frequently expressed in patients with LVNC, but without sufficient power to be used as a discriminative diagnostic parameter. Both LGE and eccentricity of the left ventricle were found to be relatively solid diagnostic landmarks of complex infrastructural and functional changes within the failing heart.

     

Pseudo‐real‐time retinal layer segmentation for high‐resolution adaptive optics optical coherence tomography

We present a pseudo‐real‐time retinal layer segmentation for high‐resolution Sensorless Adaptive Optics‐Optical Coherence Tomography (SAO‐OCT). Our pseudo‐real‐time segmentation method is based on Dijkstra's algorithm that uses the intensity of pixels and the vertical gradient of the image to find the minimum cost in a geometric graph formulation within a limited search region. It segments six retinal layer boundaries in an iterative process according to their order of prominence. The segmentation time is strongly correlated to the number of retinal layers to be segmented. Our program permits en face images to be extracted during data acquisition to guide the depth specific focus control and depth dependent aberration correction for high‐resolution SAO‐OCT systems. The average processing times for our entire pipeline for segmenting six layers in a retinal B‐scan of 496 × 400 and 240 × 400 pixels are around 25.60 and 13.76 ms, respectively. When reducing the number of layers segmented to only two layers, the time required for a 240 × 400 pixel image is 8.26 ms.