Quantification of anatomical aortic valve area by multi-detector computed tomography: A pilot 3D-morphological modeling of the stenotic aortic valve

Background: Aortic-valve-stenosis (AS) is a frequent degenerative valvular-disease and carries dismal outcome under-medical-treatment. Transvalvular pressure gradient reflects severity of the valve-disease but is highly dependent on flow-conditions and on other valvular/aortic characteristics. Alternatively, aortic-valve-area (AVA) represents a measure of aortic-valve lesion severity conceptually essential and practically widely-recognized but exhibits multiple-limitations.

Methods: We analyzed the 4D multi-detector computed tomography(MDCT) of 20 randomly selected patients with severe AS. For each-patient, we generated the 3D-model of the valve and of its calcifications, and we computed the anatomical AVA accounting for the 3D-morphology of the leaflets in three-different-ways. Finally, we compared our results vs. Doppler-based AVA_E measurements and vs. 2D-planimetric AVA-measurements.

Results: 3D-reconstruction and identification of the cusps were successful in 90% of the cases. The calcification patterns where highly-variable over patients, ranging from multiple small deposits to wide and c-shaped deposits running from commissure-to-commissure. AVA_E was 82 ± 15 mm². When segmenting 18 image planes, AVA_Tight, AVA_Proj-Ann, AVA_Proj-Tip and their average AVA_Ave were equal to 80 ± 16, 88 ± 20, 93 ± 21 and 87 ± 19 mm², respectively, while AVA_Plan was equal to 143 ± 50 mm². Linear-regression of the three measurements vs. AVA_E yielded regression slopes equal to 1.26, 1.13 and 0.93 for AVA_Proj-Ann, AVA_Proj-Tip and AVA_Tight, respectively. The respective Pearson-coefficients were 0.85,0.86 and 0.90. Conversely, when comparing AVA_Plan vs. AVA_E, linear regression yielded a slope of 1.73 and a Pearson coefficient of 0.53.

Conclusions: We described a new-method to obtain a set of flow-independent quantifications that complement pressure gradient measurements and combine the advantages of previously proposed methods, while bypassing the corresponding-limitations.