Tortuosity in tumours: The need of combining multi-phase flows with machine learning tools

This article presents a preliminary approximation of the effect in blood flow produced by tortuosity in the blood vessels. This roughness is related to certain diseases (such as tumours) and the modelling of the flow properties can help to characterize the existence of such diseases. Malignancy provokes local changes to vessel shape and characteristic vessel tortuosity appears early during tumour development. Numerical simulations have been carried out, where the elastic behaviour modelling the surface of blood vessels has been represented using several geometries with different wavy shapes. The complexity of the topology patterns, which could be related with the different particle paths found in different geometries, encourages to continue this research using novel machine learning tools to predict and connect the main flow instabilities with the tumour evolution. •Multiphase flow to simulate tortuosity in blood vessels.•Topology patterns in blood vessels with different configurations.•New flow instabilities due to the interaction of particles in channels with wavy walls.