Insights from 3D modeling and fluid dynamics in COVID-19 pneumonia.

We address the lack of research regarding aerodynamic events behind respiratory distress at COVID-19. The use of chest CT enables quantification of pneumonia extent; however, there is a paucity of data regarding the impact of airflow changes. We reviewed 31 COVID-19 patients who were admitted in March 2020 with varying severity of pulmonary disease. Lung volumes were segmented and measured on CT images and patient-specific models of the lungs were created. Incompressible, laminar, and three-dimensional Navier-Stokes equations were used for the fluid dynamics (CFD) analyses of ten patients (five mild, five pneumonia). Of 31 patients, 17 were female, 18 had pneumonia, and 2 were deceased. Effective lung volume decreased in the general group, but the involvement of the right lung was prominent in dyspnea patients. CFD analyses revealed that the mass flow distribution was significantly distorted in pneumonia cases with diminished flow rate towards the right lung. In addition, the distribution of flow parameters showed mild group had less airway resistance with higher velocity (1.228 m/s vs 1.572 m/s) and higher static pressure values at airway branches (1.5112 Pa vs 1.3024 Pa). Therefore, we conclude that airway resistance and mass flow rate distribution are as important as the radiological involvement degree in defining the disease severity.

Effect of glottic geometry on breathing: three-dimensional unsteady numerical simulation of respiration in a case with congenital glottic web.

Glottic obstruction is a major cause of dyspnea. Without understanding the normal function of the glottis in breathing, treating dyspnea does not restore normal physiology. Therefore, we designed a computational fluid dynamics (CFD) model that tested the respiratory cycle in larynges with normal glottis and congenital glottic web (CGW). A CGW case and a control subject (CC) were selected from the computed tomography (CT) archive. 3D computational models of the larynges with structured boundary layer were constructed from axial CT images after mesh refinement study. CFD analyses were based on the Reynolds-averaged Navier-Stokes approach. Incompressible flow solver (pressure-based) and SST k-w turbulence model were chosen for this study. To simulate a real-time breathing process, time varying flow rate boundary condition was derived from the spirometer of a healthy, non-smoking woman. Glottic areas were measured as 51.64 and 125.43 mm(2) for the CGW patient and CC, respectively. Time-dependent velocity contours and streamlines for the CC and CGW patient were drawn. The CC showed uniform flow, all through the inspiration and expiration phases. However, the CGW patient showed separation of flow at the glottis level, which caused areas of stagnation in the supraglottis (during expiration) and the subglottis and trachea (during inspiration). Specialized geometry of the normal larynx maintained uniform flow with low shear stress values on the wall even at high mass flow rates. Distortion of this geometry may cause obstruction of flow at multiple levels and, therefore, should be evaluated at multiple levels.

A computational study on the characteristics of airflow in bilateral abductor vocal fold immobility.

OBJECTIVES/HYPOTHESIS: To evaluate airway sufficiency and airflow dynamics in a group of patients who underwent a posterior transverse laser cordotomy (PTLC) procedure. STUDY DESIGN: Mixed methods research, university hospital setting. METHODS: Sixteen patients who underwent a PTLC procedure volunteered to be involved in this study. Dyspnea levels, voice, and glottic opening in indirect laryngoscopy were evaluated subjectively. The airway was evaluated objectively by pulmonary function tests, and glottic areas were measured from axial computed tomography (CT) images. The control group consisted of 63 subjects from the tomography archive. For computational fluid dynamics (CFD) analyses, two subjects from the study group were chosen on the basis of obstruction level, and a normal female subject was selected from the control group. Cartesian coordinates for airway boundaries were determined from axial CT images, and a three-dimensional computational model of the larynx was constructed. Flow simulations were performed with two different flow conditions during inspiration. Comparison of velocity, static pressure, turbulence intensity, and wall shear stress distribution values were made between selected cases and control. RESULTS: Pulmonary data varied widely and did not correlate with the size of the glottic area or dyspnea level. CFD analyses revealed that in addition to obstruction at the glottic level, aerodynamic properties of the larynx are altered due to loss in muscular tonus. Also, the contour of the glottic opening was found to be very important in determining the character of airflow as laminar or turbulent. CONCLUSIONS: Patients have considerable differences in their flow patterns and force distributions during respiration. Patient-specific models may help in evaluation and treatment planning.