MEGA: a computational framework to simulate the acute respiratory distress syndrome.

Abstract

The acute respiratory distress syndrome (ARDS) is a critical condition that necessitates mechanical ventilation (MV) to ensure sufficient ventilation and oxygenation for patients. Intensivists employ various therapeutic tools such as adjusting positive end-expiratory pressure (PEEP) levels or positioning the patient prone. However, practitioners encounter several challenges when dealing with ARDS: high variability among patients and limited understanding of underlying mechanisms. As a result, decision-making by physicians largely relies on experience. Yet, having the ability to estimate the likelihood of a patient responding to different therapeutic approaches would hold significant clinical value. Moreover, gaining a deeper understanding of the biomechanical and physiological phenomena underlying patient responses could inform the development of new MV strategies for ARDS management. To address these challenges, a coupled physiomechanical computational framework based on patient's computed tomography scan data was conceived and implemented. Simulations were conducted for prone positioning and PEEP-increment scenarios. The model results qualitatively align with both literature data and clinical measurements. However, some results diverge quantitatively from clinical measurements, emphasizing the necessity for thorough model calibration. Nonetheless, this serves as a proof of concept that the developed framework could be valuable in supporting intensivists' decision-making processes.NEW & NOTEWORTHY An original computational framework has been developed to simulate respiratory biomechanics and physiology of patients with ARDS. Using patient's CT scans, this spatially resolved model enables the calculation of global parameters (e.g., tidal volumes), but also the detailed distribution of ventilation within the lung, a capability not achievable with conventional single-compartment models commonly used in clinical practice. Furthermore, the framework allows to simulate recruitment maneuvers, including those regularly performed in ICU, such as prone positioning.