A Preprocessing Approach for Innovative Patient-specific Intranasal Flow Simulations

Abstract

A preprocessing approach which enables numerical simulations of patient-individual respiration flows based on computer tomography (CT) scans is presented. The challenge of an adequate 3D reconstruction and preparation of as highly complex geometries as the human inner nose or lung, which cannot be resolved in all their details by nowadays available CT scanners, is met by a holistic concept. In it, the preprocessing is understood as one part of the full numerical simulation where preprocessing, numerical simulation and optimization as well as postprocessing techniques are applied in an iteratively coupled fashion. In the proposed approach, lattice Boltzmann methods are chosen as discretization strategy in order to simulate the airflows. Since they impose the conditions for the preparation process of the geometries, this choice is of crucial importance for the preprocessing techniques which are to be applied. For the actual preprocessing steps, it is proposed to take advantage of techniques, which are implemented in the framework of Materialise's software packages Mimics and 3-matics. In this paper, main emphasis is placed on illustrating the preprocessing approach as part of the proposed full numerical simulation concept. Therefore, the whole chain of necessary processes from the segmentation over the surface generation, the actual volume mesh generation, the numerical simulation, the validation up to the visualization of the results is demonstrated by means of an example, namely the full numerical simulation of an expiration in an inner human nose. The researched geometry belongs to a patient with a severe peripheral obstructive ventilation disorder. Based on the obtained numerical results, possible causes and consequences are discussed, in particular, a stenosis is located.