Numerical Study on the Impact of Central Venous Catheter Placement on Blood Flow in the Cavo-Atrial Junction.

An in silico study is performed to investigate fluid dynamic effects of central venous catheter (CVC) placement within patient-specific cavo-atrial junctions. Prior studies show the CVC infusing a liquid, but this study focuses on the placement without any liquid emerging from the CVC. A 7 or 15-French double-lumen CVC is placed virtually in two patient-specific models; the CVC tip location is altered to understand its effect on the venous flow field. Results show that the CVC impact is trivial on flow in the superior vena cava when the catheter-to-vein ratio ranges from 0.15 to 0.33. Results further demonstrate that when the CVC tip is directly in the right atrium, flow vortices in the right atrium result in elevated wall shear stress near the tip hole. A recirculation region characterizes a spatially variable flow field inside the CVC side hole. Furthermore, flow stagnation is present near the internal side hole corners but an elevated wall shear stress near the curvature of the side hole's exit. These results suggest that optimal CVC tip location is within the superior vena cava, so as to lower the potential for platelet activation due to elevated shear stresses and that CVC geometry and location depth in the central vein significantly influences the local CVC fluid dynamics. A thrombosis model also shows thrombus formation at the side hole and tip hole. After modifying the catheter design, the hemodynamics change, which alter thrombus formation. Future studies are warranted to study CVC design and placement location in an effort to minimize CVC-induced thrombosis incidence.

Data-smart machine learning methods for predicting composition-dependent Young's modulus of pharmaceutical compacts.

The ability to predict mechanical properties of compacted powder blends of Active Pharmaceutical Ingredients (API) and excipients solely from component properties can reduce the amount of 'trial-and-error' involved in formulation design. Machine Learning (ML) can reduce model development time and effort with the imperative of adequate historical data. This work describes the utility of linear and nonlinear ML models for predicting Young's modulus (YM) of directly compressed blends of known excipients and unknown API mixed at arbitrary compositions given only the true density of the API. The models were trained with data from compacts of three BCS Class I APIs and two excipients blended at four drug loadings, three excipient compositions, and compacted to five nominal solid fractions. The prediction accuracy of the models was measured using three cross-validation (CV) schemes. Finally, we demonstrate an application of the model to enable Quality-by-Design in formulation design. Limitations of the models and future work have also been discussed.