In the biofluids space, the Digital Twin paradigm relies on leveraging medical images and pressure and flow data to perform customized, subject-specific predictions. Flows in blood vessels and lymphatic systems include complex structured fluids, composed of a liquid phase (plasma) and a disperse phase, which includes cells and other particles such as thrombus. In this work, we present two novel numerical formulations for digital twinning in biofluids, implemented in the open-source framework CRIMSON [1]. In the first application, we are concerned with the study of conditions in which the complex interactions between plasma and suspended particles is required. We recently developed a volume-filtered Eulerian-Lagrangian strategy that uses a finite element method (FEM) to solve for the fluid phase coupled with a discrete element method (DEM) for the particle phase [2]. In a second application, we are concerned with the diagnosis of vascular disease in the catheterization laboratory. Here, images and other hemodynamic data are acquired in real time, and the cardiologist must decide in just a few minutes whether the blockage in the vessels is severe enough to warrant intervention. This application relies on computational tools to accurately estimate pressure gradients and other parameters within minutes. This goal can only be achieved via suitable reduced order models. Our group has recently developed a machine learning approach for calculating pressure gradient indices in image-based vascular models [3]. REFERENCES [1] C. Arthurs, et al. “CRIMSON: An open-source software framework for cardiovascular integrated modelling and simulation”. In: PLOS Computational Biology 17.5 (May 2021), pp. 1–21. doi:10.1371/journal.pcbi.1008881. [2] Abhilash Reddy Malipeddi, C. Alberto Figueroa, and Jesse Capecelatro. Volume filtered FEM-DEM framework for simulating particle-laden flows in complex geometries. Nov. 2023. eprint: arXiv:2311.15989. http://arxiv.org/abs/2311.15989. [3] Haizhou Yang, C. Alberto Figueroa, and Krishna Garikipati. Attention-based Multi-fidelity Machine Learning Model for Computational Fractional Flow Reserve Assessment. 2023. arXiv: 2311.11397