Ching-Long Lin

Abstract In this chapter, we review the current state of in silico tools for lung deposition applications and discuss how knowledge gathered from recent studies can be purposefully leveraged to design efficient hybrid multiscale lung models. We firstly address the different roles of in silico methods when applied to the human lung airways. We then discuss the variability of airflow and particle transport regimes in different regions of the human lung and how in silico methods are employed in each case. In the next section, we focus on numerical aspects associated with the application of computational fluid particle dynamics (CFPD) in the extrathoracic and upper conducting airways. We highlight the variability in airflow and deposition predictions in the upper airways across different CFPD methods and the necessity for validation and verification of computational tools. A review of models applied to the pulmonary acinus follows, in which we highlight how CFPD methods can deliver high-resolution spatiotemporal predictions of local acinar aerosol deposition. Towards integrated simulations covering the whole lung, we describe efforts to integrate 3D CFPD for the upper airways with models of the peripheral lung and discuss open issues and expected developments. We then discuss how CFPD can account for the non-standard lung, namely disease, age and gender effects. We close with a brief review of 1D models and how their predictions can be improved in the future by leveraging knowledge generated by the more complex 3D CFPD and hybrid approaches.

Abstract This chapter reviews the techniques and strategies for identifying subpopulations (clusters) characterized by distinct lung features via machine learning and using cluster information to guide in silico computational fluid and particle dynamics (CFPD) analysis for the design of future inhaled drug delivery methods. We first review the collaborative efforts of collecting imaging, genetic, clinical and biological data sets for large cohorts of healthy, asthma and chronic obstructive pulmonary disease (COPD) subjects to investigate the heterogeneous nature of lung disease. We then focus on imaging-based phenotyping due to its quantitative nature that sensitively captures lung structural and functional alternations at both local (segmental/parenchymal) and global (lobar/lung) scales. Machine learning is then applied to identify imaging clusters for asthma and COPD patients. We select cluster archetypes to perform CFPD analysis and use CFPD-derived variables to interpret the link between cluster-specific alterations and particle depositions in the human lungs. Finally, we discuss the prospect of employing machine learning, physics-based learning and deep learning complementarily toward precision medicine.

Accurate prediction of airflow distribution and aerosol transport in the human lungs, which are difficult to be measured in vivo but important to understand the structure and function relationship, is challenging. It is because the interplay between them spans more than two orders of magnitude in dimension from the trachea to alveoli. This chapter reviews the techniques and strategies for modeling lungs both within and between subjects, viz., subject specificity versus generalization from individuals to populations, with both exhibiting multiscale characteristics. For “within-subject” modeling, a computed tomography (CT)-derived subject-specific computational fluid dynamics (CFD) lung model is presented. The pipeline for building such an imaging-based lung model is composed of image segmentation and processing, geometrical modeling labeled with anatomical information, image registration, three-dimensional (accurate) and one-dimensional (approximate) coupling techniques, and a high-fidelity turbulent flow model. The subject-specific model is essential in predicting local structural and functional interactions. For “between subjects” modeling, machine learning is employed to identify homogeneous subpopulations (clusters), among healthy and diseased populations, aiming to bridge individual and population scales. For this purpose, three major issues need to be addressed. They are intersubject variability (due to, e.g., gender, age, and height), inter-site variability (due to scanner and imaging protocol differences), and definition of quantitative CT imaging-based metrics at multiple scales (due to alterations at different disease stages) needed for clustering. The use of the cluster membership to select representative subjects for detailed CFD analysis enables an examination of the cluster-specific structural and functional relationships.

The human lungs are divided into five independent compartments called lobes. The lobar fissures separate the lung lobes. It is hypothesized that the lobar surfaces slide against each other during respiration. We propose a method to evaluate the sliding motion of the lobar surfaces during respiration using lobe-by-lobe mass-preserving non-rigid image registration. We measure lobar sliding by evaluating the relative displacement on both sides of the fissure. The results show a superior-inferior gradient in the magnitude of lobar sliding. We compare whole-lung-based registration accuracy to lobe-by-lobe registration accuracy using vessel bifurcation landmarks.

Basic exploration of diffusion equation solvers in distributed computing systems has been a very important issue for computational fluid dynamics (CFD). This paper presents a fundamental study of a distributed computing solution for diffusive phenomena in a P2P enhanced Grid system. We propose a simple distributed system with the architecture of multi-clients and multi-servers (MCMS). Multithreading is implemented on both the client and server nodes, enabling them to communicate. This paper focuses on architecture, model implementation, multithreading, network communication, domain decomposition/composition, problem convergence, fault-tolerance, and dynamic load balancing. A discussion of the results from conducted numerical experiments is useful for performing future intensive CFD computations on Grids.

In this paper we examine the interaction of internal waves with the critical layer and the role of the parametric subharmonic instability (PSI) in the energy transfer process when multiple triad interactions are generated by one strong primary mode and background white noise or a Garrett‐Munk (GM79) model spectrum. We performed direct numerical simulations of the incompressible Navier‐Stokes equations with the Boussinesq approximation, using the pseudo‐spectral method (see Rogallo, 1981, Holt et al., 1992). For the critical layer a spanwise Rayleigh‐Taylor instability should be the most unstable mode if the Richardson number is less than zero locally. The mixing efficiency associated with this process still remains an open question. For the triad interactions the PSI energy transfer mechanism is feasible if local sum resonance is suppressed (i.e. under certain geometric constraints), but it is not clear if the mechanism is resonant. When examining triads with a background GM79 spectrum we found that even when the energy level of the spectrum was reduced by a factor of 100 the induced diffusion triads (non‐resonant) were the strongest, indicating the dominance of wave‐mean flow interactions. In this instance triads of the PSI type were still not found, suggesting that they may not be significant in the distribution of wave energy in the ocean.