Development and progression of cardiovascular diseases are closely linked to patient-specific alterations in flow and hemodynamics, making restoration of physiological conditions a recognized therapeutic target. However, healthy flow characteristics are typically unknown at the individual level. Thus, simplified or generic healthy inflow profiles commonly used in computational models may yield misleading results. This study proposes a novel data-driven framework to generate patient-specific synthetic healthy inflow profiles for aortic valve stenosis (AS) patients. Using space- and time-resolved 4D Flow MRI data over the cardiac cycle from N = 13 healthy volunteers, statistical shape modeling and principal component analysis were employed to derive synthetic 3D supravalvular flow profiles capturing the dominant features in healthy aortic flow. These synthetic profiles were scaled and adjusted using subject-specific flow descriptors, e.g. flow dispersion index, flow jet angle, secondary flow degree. Two additional healthy datasets were used for validation. The framework was subsequently applied to 4D Flow MRI data from n = 5 AS patients to generate individualized synthetic healthy reference flows, which were evaluated against established healthy flow characteristics. The approach accurately reproduced healthy validation flows with deviations in flow descriptors below 10%. For AS cohort, the pipeline allows a patient-specific generation of healthy synthetic references given individual pathological inflow data, yielding physiologically realistic, time-resolved flow structures consistent with healthy aortic hemodynamics. Future work will focus on extending the framework to larger and more diverse populations and diseases, as well as on integrating the synthetic profiles as inflow boundary conditions for computational fluid dynamics simulations to support diagnosis, therapy