Three-dimensional cardiac computed tomography compared with autopsied material for the assessment of the mitral valve.
Agata Krawczyk-OżógMateusz Krystian HołdaJakub BatkoStanisław BartuśRenata Rajtar-SalwaPublished in: Clinical anatomy (New York, N.Y.) (2022)
To compare the morphometrical features of non-diseased mitral valves imaged in three-dimensional (3D) cardiac computed tomography with those analyzed macroscopically in autopsied healthy human hearts. A total of 51 cardiac computed tomography scans and 120 adult autopsied human hearts without cardiovascular disease were examined. The 3D reconstruction and visualization software (Mimics Innovation Suite 22, Materialise) was used for heart chambers semi-automatic segmentation and myocardial manual segmentation to visualize a 3D structure of the mitral valve complex and to perform all measurements. Direct comparison of corresponding mitral valve parameters revealed significant differences between obtained results. Significantly larger intercommisural diameter, aorto-mural diameter, and perimeter of the mitral annulus were observed in tomographic scans (all p < 0.0001). However, the intercommissural/aorto-mural diameter ratio showed comparable values for both groups. Nevertheless, the size of anterior mitral leaflet was higher in autopsied material. The height of the P2 scallops was the only parameter that show no significant difference between two groups (p = 0.3). The use of 3D postprocessing algorithms provides a very accurate image of the mitral valve structure, which could be useful for the precise non-invasive assessment of mitral valve size and structure. Three-dimensional contrast enhanced cardiac computed tomography significantly overestimates the measurements of the mitral annulus compared to postmortem analysis.
Keyphrases
- mitral valve
- computed tomography
- left ventricular
- contrast enhanced
- deep learning
- positron emission tomography
- left atrial
- magnetic resonance imaging
- dual energy
- cardiovascular disease
- heart failure
- endothelial cells
- diffusion weighted
- aortic valve
- magnetic resonance
- aortic stenosis
- image quality
- convolutional neural network
- optic nerve
- high resolution
- mass spectrometry
- type diabetes
- body mass index
- metabolic syndrome
- induced pluripotent stem cells
- single cell
- physical activity