Bicuspid aortic valve (BAV) repair is a promising surgical treatment for young adults with aortic regurgitation (AR). However, BAV repair surgery remains underutilized and variably applied across institutions, owing in part to the lack of a standardized approach to BAV repair planning. Currently, BAV repair planning relies primarily on intraoperative manual measurements of the valve made by direct observation while the heart is in an arrested state, making it difficult for the surgeon to identify defects in valve dynamics under physiological conditions. To address this challenge, the long-term goal is to develop a multimodal 4D image analytics and valve modeling platform that systematically characterizes pre-operative BAV morphology and dynamics and enables patient-specific surgical planning. The overall objectives of this proposal are to (i) fill a knowledge gap in the precise anatomical relationships between the aortic cusps, annulus, and root that make a BAV functionally competent, and (ii) develop computational image analytics to precisely identify the patient- specific, anatomical and dynamic distortions that cause AR so that these defects can be prioritized for risk stratification and planning of BAV repair surgery. This work will be carried out by pursuing three specific aims: (1) Design and assess an automated segmentation and modeling algorithm for 4D reconstruction of the BAV apparatus from multiple clinical imaging modalities; (2) characterize the morphological and dynamic features of BAV competence and create a machine learning method for comprehensive anomaly detection in regurgitant BAVs; (3) evaluate a BAV repair planning system using images acquired from valve repair procedures at three institutions. The proposed project leverages the complementary benefits of two modalities: real-time 3D transesophageal echocardiography and 4D computed tomography angiography, which capture both the morphological detail of the aortic cusps with high spatial resolution and the motion of the 3D BAV apparatus with high temporal resolution. The innovation of this project is that the proposed tools could change how BAV repair planning is carried out. Instead of relying on intraoperative inspection of the valve while it is unpressurized, the surgeon can interactively visualize image-derived BAV models and quantify dynamic mechanisms of AR when the valve is in a pre-operative 4D physiological state. The significance of this research is that it could promote consistency in valve repair planning across institutions, decrease surgeons’ reliance on intuition and trial-and-error, and thereby increase the utilization of BAV repair in young adults. This would have quality of life advantages relative to conventional valve replacement, which requires lifelong anticoagulation therapy (mechanical valves) or multiple re-replacements due to limited durability (bioprosthetic valves). Ultimately, the systematic analysis of multimodal image data for computer-aided valve...