Pancreas image mining: a systematic review of radiomics.

OBJECTIVES: To systematically review published studies on the use of radiomics of the pancreas. METHODS: The search was conducted in the MEDLINE database. Human studies that investigated the applications of radiomics in diseases of the pancreas were included. The radiomics quality score was calculated for each included study. RESULTS: A total of 72 studies encompassing 8863 participants were included. Of them, 66 investigated focal pancreatic lesions (pancreatic cancer, precancerous lesions, or benign lesions); 4, pancreatitis; and 2, diabetes mellitus. The principal applications of radiomics were differential diagnosis between various types of focal pancreatic lesions (n = 19), classification of pancreatic diseases (n = 23), and prediction of prognosis or treatment response (n = 30). Second-order texture features were most useful for the purpose of differential diagnosis of diseases of the pancreas (with 100% of studies investigating them found a statistically significant feature), whereas filtered image features were most useful for the purpose of classification of diseases of the pancreas and prediction of diseases of the pancreas (with 100% of studies investigating them found a statistically significant feature). The median radiomics quality score of the included studies was 28%, with the interquartile range of 22% to 36%. The radiomics quality score was significantly correlated with the number of extracted radiomics features (r = 0.52, p < 0.001) and the study sample size (r = 0.34, p = 0.003). CONCLUSIONS: Radiomics of the pancreas holds promise as a quantitative imaging biomarker of both focal pancreatic lesions and diffuse changes of the pancreas. The usefulness of radiomics features may vary depending on the purpose of their application. Standardisation of image acquisition protocols and image pre-processing is warranted prior to considering the use of radiomics of the pancreas in routine clinical practice. KEY POINTS: • Methodologically sound studies on radiomics of the pancreas are characterised by a large sample size and a large number of extracted features. • Optimisation of the radiomics pipeline will increase the clinical utility of mineable pancreas imaging data. • Radiomics of the pancreas is a promising personalised medicine tool in diseases of the pancreas.

Evaluating the role of sex-related structure-function differences on airway aerosol transport and deposition.

Several experimental studies have found that females have higher deposition of particles in the airways compared with males. This has implications for the delivery of aerosolized therapeutics and for understanding sex differences in respiratory system response to environmental exposures. This study evaluates several factors that potentially contribute to sex differences in particle deposition, using scale-specific structure-function models of 1D ventilation distribution, particle transport, and deposition. The impact of gravity, inhalation flow rate, and dead space are evaluated in 12 structure-based models (seven female; five male). Females were found to have significantly higher total, bronchial, and alveolar deposition than males across a particle size range from 0.01 to 10 . Results suggest that higher deposition fraction in females is due to higher alveolar deposition for smaller particle sizes, and higher bronchial deposition for larger particles. Females had higher alveolar deposition in the lower lobes, and slightly lower particle concentration in the left upper lobe. Males were found to be more sensitive to changes due to gravity, showing greater reduction in bronchial deposition fraction. Males were also more sensitive to change in inhalation flow rate, and to scaling of dead space due to the larger male baseline airway size. Predictions of sex differences in particle deposition - that are consistent with the literature - suggest that sex-based characteristics of lung and airway size interacting with particle size gives rise to differences in regional deposition.

3D Quantitative-Amplified Magnetic Resonance Imaging (3D q-aMRI).

Amplified MRI (aMRI) is a promising new technique that can visualize pulsatile brain tissue motion by amplifying sub-voxel motion in cine MRI data, but it lacks the ability to quantify the sub-voxel motion field in physical units. Here, we introduce a novel post-processing algorithm called 3D quantitative amplified MRI (3D q-aMRI). This algorithm enables the visualization and quantification of pulsatile brain motion. 3D q-aMRI was validated and optimized on a 3D digital phantom and was applied in vivo on healthy volunteers for its ability to accurately measure brain parenchyma and CSF voxel displacement. Simulation results show that 3D q-aMRI can accurately quantify sub-voxel motions in the order of 0.01 of a voxel size. The algorithm hyperparameters were optimized and tested on in vivo data. The repeatability and reproducibility of 3D q-aMRI were shown on six healthy volunteers. The voxel displacement field extracted by 3D q-aMRI is highly correlated with the displacement measurements estimated by phase contrast (PC) MRI. In addition, the voxel displacement profile through the cerebral aqueduct resembled the CSF flow profile reported in previous literature. Differences in brain motion was observed in patients with dementia compared with age-matched healthy controls. In summary, 3D q-aMRI is a promising new technique that can both visualize and quantify pulsatile brain motion. Its ability to accurately quantify sub-voxel motion in physical units holds potential for the assessment of pulsatile brain motion as well as the indirect assessment of CSF homeostasis. While further research is warranted, 3D q-aMRI may provide important diagnostic information for neurological disorders such as Alzheimer's disease.

A Wearable Open-Source electrical impedance tomography device.

Electrical impedance tomography (EIT) is medical imaging technique in which small electrical signals are used to map the electrical impedance distribution within the body. It is safe and non-invasive, which make it attractive for use in continuous monitoring or outpatient applications, but the high cost of commercial devices is an impediment to its adoption. Over the last 10 years, many research groups have developed their own EIT devices, but few designs for open-source EIT hardware are available. In this work, we present a complete open-source EIT system that is designed to be suitable for monitoring the lungs of free breathing subjects. The device is low-cost, wearable, and is designed to comply with the industry accepted safety standard for EIT. The device has been tested in two regimes: Firstly in terms of measurement uncertainty as a voltage measurement system, and secondly against a set of measures that have been proposed specifically for EIT hardware. The voltage measurement uncertainty of the device was measured to be - 0.7 % ± 0.36 mV. The EIT specific performance was measured in a phantom test designed to be as physiologically representative as practicable, and the device performed similarly to other published devices. This work will contribute to increased accessibility of EIT for study and will contribute to consensus on testing methodology for EIT devices.

Evaluating Tissue Heterogeneity in the Radiologically Normal-Appearing Tissue in IPF Compared to Healthy Controls.

RATIONALE AND OBJECTIVES: Idiopathic Pulmonary Fibrosis (IPF) is a progressive interstitial lung disease characterised by heterogeneously distributed fibrotic lesions. The inter- and intra-patient heterogeneity of the disease has meant that useful biomarkers of severity and progression have been elusive. Previous quantitative computed tomography (CT) based studies have focussed on characterising the pathological tissue. However, we hypothesised that the remaining lung tissue, which appears radiologically normal, may show important differences from controls in tissue characteristics. MATERIALS AND METHODS: Quantitative metrics were derived from CT scans in IPF patients (N = 20) and healthy controls with a similar age (N = 59). An automated quantitative software (CALIPER, Computer-Aided Lung Informatics for Pathology Evaluation and Rating) was used to classify tissue as normal-appearing, fibrosis, or low attenuation area. Densitometry metrics were calculated for all lung tissue and for only the normal-appearing tissue. Heterogeneity of lung tissue density was quantified as coefficient of variation and by quadtree. Associations between measured lung function and quantitative metrics were assessed and compared between the two cohorts. RESULTS: All metrics were significantly different between controls and IPF (p < 0.05), including when only the normal tissue was evaluated (p < 0.04). Density in the normal tissue was 14% higher in the IPF participants than controls (p < 0.001). The normal-appearing tissue in IPF had heterogeneity metrics that exhibited significant positive relationships with the percent predicted diffusion capacity for carbon monoxide. CONCLUSION: We provide quantitative assessment of IPF lung tissue characteristics compared to a healthy control group of similar age. Tissue that appears visually normal in IPF exhibits subtle but quantifiable differences that are associated with lung function and gas exchange.

Quantitative Analysis of Lung Shape in Idiopathic Pulmonary Fibrosis: Insights Into Disease- and Age-Associated Patterns.

RATIONALE AND OBJECTIVES: Fibrotic scarring in idiopathic pulmonary fibrosis (IPF) typically develops first in the posterior-basal lung tissue before advancing to involve more of the lung. The complexity of lung shape in the costo-diaphragmatic region has been proposed as a potential factor in this regional development. Intrinsic and disease-related shape could therefore be important for understanding IPF risk and its staging. We hypothesized that lung and lobe shape in IPF would have important differences from controls. MATERIALS AND METHODS: A principal component (PC) analysis was used to derive a statistical shape model (SSM) of the lung for a control cohort aged > 50 years (N = 39), using segmented lung and fissure surface data from CT imaging. Individual patient shape models derived for baseline (N = 18) and follow-up (N = 16) CT scans in patients with IPF were projected to the SSM to describe shape as the sum of the SSM average and weighted PC modes. Associations between the first four PC shape modes, lung function, percentage of fibrosis (fibrosis%) and pulmonary vessel-related structures (PVRS%), and other tissue metrics were assessed and compared between the two cohorts. RESULTS: Shape was different between IPF and controls (P < 0.05 for all shape modes), with IPF shape forming a distinct shape cluster. Shape had a negative relationship with age in controls (P = 0.013), but a positive relationship with age in IPF (P = 0.026). Some features of shape changed on follow-up. Shape in IPF was associated with fibrosis% (P < 0.05) and PVRS% (P < 0.05). CONCLUSION: Quantitative comparison of lung and lobe shape in IPF with controls of a similar age reveals shape differences that are strongly associated with age and percent fibrosis. The clustering of IPF cohort shape suggests that it could be an important feature to describe disease.

Improved Electrical Impedance Tomography Reconstruction via a Bayesian Approach With an Anatomical Statistical Shape Model.

OBJECTIVE: electrical impedance tomography (EIT) is a promising technique for rapid and continuous bedside monitoring of lung function. Accurate and reliable EIT reconstruction of ventilation requires patient-specific shape information. However, this shape information is often not available and current EIT reconstruction methods typically have limited spatial fidelity. This study sought to develop a statistical shape model (SSM) of the torso and lungs and evaluate whether patient-specific predictions of torso and lung shape could enhance EIT reconstructions in a Bayesian framework. METHODS: torso and lung finite element surface meshes were fitted to computed tomography data from 81 participants, and a SSM was generated using principal component analysis and regression analyses. Predicted shapes were implemented in a Bayesian EIT framework and were quantitatively compared to generic reconstruction methods. RESULTS: Five principal shape modes explained 38% of the cohort variance in lung and torso geometry, and regression analysis yielded nine total anthropometrics and pulmonary function metrics that significantly predicted these shape modes. Incorporation of SSM-derived structural information enhanced the accuracy and reliability of the EIT reconstruction as compared to generic reconstructions, demonstrated by reduced relative error, total variation, and Mahalanobis distance. CONCLUSION: As compared to deterministic approaches, Bayesian EIT afforded more reliable quantitative and visual interpretation of the reconstructed ventilation distribution. However, no conclusive improvement of reconstruction performance using patient specific structural information was observed as compared to the mean shape of the SSM. SIGNIFICANCE: The presented Bayesian framework builds towards a more accurate and reliable method for ventilation monitoring via EIT.

De novo discovery of traits co-occurring with chronic obstructive pulmonary disease.

Chronic obstructive pulmonary disease (COPD) is a heterogeneous group of chronic lung conditions. Genome-wide association studies have identified single-nucleotide polymorphisms (SNPs) associated with COPD and the co-occurring conditions, suggesting common biological mechanisms underlying COPD and these co-occurring conditions. To identify them, we have integrated information across different biological levels (i.e., genetic variants, lung-specific 3D genome structure, gene expression and protein-protein interactions) to build lung-specific gene regulatory and protein-protein interaction networks. We have queried these networks using disease-associated SNPs for COPD, unipolar depression and coronary artery disease. COPD-associated SNPs can control genes involved in the regulation of lung or pulmonary function, asthma, brain region volumes, cortical surface area, depressed affect, neuroticism, Parkinson's disease, white matter microstructure and smoking behaviour. We describe the regulatory connections, genes and biochemical pathways that underlay these co-occurring trait-SNP-gene associations. Collectively, our findings provide new avenues for the investigation of the underlying biology and diverse clinical presentations of COPD. In so doing, we identify a collection of genetic variants and genes that may aid COPD patient stratification and treatment.

Pulmonary vessel volume in idiopathic pulmonary fibrosis compared with healthy controls aged > 50 years.

Idiopathic pulmonary fibrosis (IPF) is characterised by progressive fibrosing interstitial pneumonia with an associated irreversible decline in lung function and quality of life. IPF prevalence increases with age, appearing most frequently in patients aged > 50 years. Pulmonary vessel-like volume (PVV) has been found to be an independent predictor of mortality in IPF and other interstitial lung diseases, however its estimation can be impacted by artefacts associated with image segmentation methods and can be confounded by adjacent fibrosis. This study compares PVV in IPF patients (N = 21) with PVV from a healthy cohort aged > 50 years (N = 59). The analysis includes a connected graph-based approach that aims to minimise artefacts contributing to calculation of PVV. We show that despite a relatively low extent of fibrosis in the IPF cohort (20% of the lung volume), PVV is 2-3 times higher than in controls. This suggests that a standardised method to calculate PVV that accounts for tree connectivity could provide a promising tool to provide early diagnostic or prognostic information in IPF patients and other interstitial lung disease.

Roadmap for an imaging and modelling paediatric study in rural NZ.

Our study methodology is motivated from three disparate needs: one, imaging studies have existed in silo and study organs but not across organ systems; two, there are gaps in our understanding of paediatric structure and function; three, lack of representative data in New Zealand. Our research aims to address these issues in part, through the combination of magnetic resonance imaging, advanced image processing algorithms and computational modelling. Our study demonstrated the need to take an organ-system approach and scan multiple organs on the same child. We have pilot tested an imaging protocol to be minimally disruptive to the children and demonstrated state-of-the-art image processing and personalized computational models using the imaging data. Our imaging protocol spans brain, lungs, heart, muscle, bones, abdominal and vascular systems. Our initial set of results demonstrated child-specific measurements on one dataset. This work is novel and interesting as we have run multiple computational physiology workflows to generate personalized computational models. Our proposed work is the first step towards achieving the integration of imaging and modelling improving our understanding of the human body in paediatric health and disease.

Simulating Multi-Scale Pulmonary Vascular Function by Coupling Computational Fluid Dynamics With an Anatomic Network Model.

The function of the pulmonary circulation is truly multi-scale, with blood transported through vessels from centimeter to micron scale. There are scale-dependent mechanisms that govern the flow in the pulmonary vascular system. However, very few computational models of pulmonary hemodynamics capture the physics of pulmonary perfusion across the spatial scales of functional importance in the lung. Here we present a multi-scale model that incorporates the 3-dimensional (3D) complexities of pulmonary blood flow in the major vessels, coupled to an anatomically-based vascular network model incorporating the multiple contributing factors to capillary perfusion, including gravity. Using the model we demonstrate how we can predict the impact of vascular remodeling and occlusion on both macro-scale functional drivers (flow distribution between lungs, and wall shear stress) and micro-scale contributors to gas exchange. The model predicts interactions between 3D and 1D models that lead to a redistribution of blood between postures, both on a macro- and a micro-scale. This allows us to estimate the effect of posture on left and right pulmonary artery wall shear stress, with predictions varying by 0.75-1.35 dyne/cm2 between postures.

Bridging the gap between respiratory research and health literacy: an interactive web-based platform.

Many patients with respiratory disease lack an understanding of basic respiratory physiology and the changes occurring in their lungs due to disease. Describing how the lungs work using realistic 3D visualisation of lung structure and function will improve communication of complicated concepts, resulting in improved health literacy. We developed a web-based platform, using anatomically realistic 3D lung models, to create an interactive visualisation tool to improve health literacy for patients with respiratory disease. A small amount of non-identifying personal information including gender, age, weight, height and smoking history can be used to customise the visualisation to an individual user. 3D computer modelling was used to create a web-based application that helps people understand how their lungs work in health and disease. The web-based application includes pages describing and visualising how the lungs work and the changes that occur during asthma and damage that smoking may be doing to their lungs. The application is freely available and located at https://sites.bioeng.auckland.ac.nz/silo6/lung_new/. This application bridges the gap between computational modelling and patient education, giving a visually compelling view into the patient's body that cannot be provided with any existing tools, hence providing a novel platform for enhancing patient-clinician interaction.

A computational model of contributors to pulmonary hypertensive disease: impacts of whole lung and focal disease distributions.

Pulmonary hypertension has multiple etiologies and so can be difficult to diagnose, prognose, and treat. Diagnosis is typically made via invasive hemodynamic measurements in the main pulmonary artery and is based on observed elevation of mean pulmonary artery pressure. This static mean pressure enables diagnosis, but does not easily allow assessment of the severity of pulmonary hypertension, nor the etiology of the disease, which may impact treatment. Assessment of the dynamic properties of pressure and flow data obtained from catheterization potentially allows more meaningful assessment of the strain on the right heart and may help to distinguish between disease phenotypes. However, mechanistic understanding of how the distribution of disease in the lung leading to pulmonary hypertension impacts the dynamics of blood flow in the main pulmonary artery and/or the pulmonary capillaries is lacking. We present a computational model of the pulmonary vasculature, parameterized to characteristic features of pulmonary arterial hypertension and chronic thromboembolic pulmonary hypertension to help understand how the two conditions differ in terms of pulmonary vascular response to disease. Our model incorporates key features known to contribute to pulmonary vascular function in health and disease, including anatomical structure and multiple contributions from gravity. The model suggests that dynamic measurements obtained from catheterization potentially distinguish between distal and proximal vasculopathy typical of pulmonary arterial hypertension and chronic thromboembolic pulmonary hypertension. However, the model suggests a non-linear relationship between these data and vascular structural changes typical of pulmonary arterial hypertension and chronic thromboembolic pulmonary hypertension which may impede analysis of these metrics to distinguish between cohorts.

Olfactory Targeting of Microparticles Through Inhalation and Bi-directional Airflow: Effect of Particle Size and Nasal Anatomy.

Background: Targeting drugs to the olfactory region in the nasal cavity can bypass the restrictive blood-brain barrier and enhance their direct delivery to the brain. However, complex nasal geometry and its demographical variations can pose challenges for targeted drug deposition in the olfactory region. Deposition of particles in the nasal cavity is influenced by particle size, airflow rate, and nasal geometry. Therefore, this study investigated the effect of these parameters on regional microparticle deposition with the view to provide insights into the nose-to-brain delivery of drugs. Methods: In this study, three anatomically accurate human nasal cavities were reconstructed in silico and deposition of microparticles under nebulization and bi-directional airflow conditions was simulated. Microparticle deposition data were analyzed to gain insight into the effect of particle size and nasal geometry. Results: Maximum olfactory deposition was observed with particles in the size range of 8 to 12 µm under nebulization and 14 to 18 µm under bi-directional airflow condition. Geometric differences between subjects were shown to significantly impact overall and regional particle deposition and introduced inter-subject variability. Significant intra-subject variability in microparticle deposition was also observed in the bi-directional delivery cases. Conclusions: The data from this study suggest that tailoring particle size, combined with a delivery protocol, may provide a unique and pragmatic way to target drugs to the olfactory region. Differences in nasal anatomy among humans can cause variability in particle deposition and need to be considered in any future applications.

Lung and fissure shape is associated with age in healthy never-smoking adults aged 20-90 years.

Lung shape could hold prognostic information for age-related diseases that affect lung tissue mechanics. We sought to quantify mean lung shape, its modes of variation, and shape associations with lung size, age, sex, and Body Mass Index (BMI) in healthy subjects across a seven-decade age span. Volumetric computed tomography from 83 subjects (49 M/34 F, BMI [Formula: see text]) was used to derive two statistical shape models using a principal component analysis. One model included, and the other controlled for, lung volume. Volume made the strongest contribution to shape when it was included. Shape had a strong relationship with age but not sex when volume was controlled for, and BMI had only a small but significant association with shape. The first principal shape mode was associated with decrease in the antero-posterior dimension from base to apex. In older subjects this was rapid and obvious, whereas younger subjects had relatively more constant dimension. A shift of the fissures of both lungs in the basal direction was apparent for the older subjects, consistent with a change in tissue elasticity with age. This study suggests a quantifiable structure-function relationship for the healthy adult lung that can potentially be exploited as a normative description against which abnormal can be compared.

Review: The role of computational simulation in understanding the postoperative sinonasal environment.

Nasal surgery improves symptoms in a majority of patients for whom medical treatment has failed. In rhinosinusitis patients, endoscopic sinus surgery aims to alleviate obstruction and re-establish mucociliary clearance. Surgery alters the structure-function relationship within the nasal passage, which is difficult to assess clinically. Computational modelling has been used to investigate this relationship by simulating air flow and environmental variables inside realistic three-dimensional models of the human nasal airway but many questions remain unanswered and need further investigation. The application of computational models to improve pre-surgical planning and post-surgical treatment may not be currently possible due to the absence of knowledge correlating the model-predicted parameters to physiological variables. Links between these parameters to patient outcomes are yet to be established. This article reviews the recent application of computational modelling to understand the nasal structure-function relationship following surgery in patients with sinusitis and nasal obstruction.

Review: The role of computational simulation in understanding the postoperative sinonasal environment.

Nasal surgery improves symptoms in a majority of patients for whom medical treatment has failed. In rhinosinusitis patients, endoscopic sinus surgery aims to alleviate obstruction and re-establish mucociliary clearance. Surgery alters the structure-function relationship within the nasal passage, which is difficult to assess clinically. Computational modelling has been used to investigate this relationship by simulating air flow and environmental variables inside realistic three-dimensional models of the human nasal airway but many questions remain unanswered and need further investigation. The application of computational models to improve pre-surgical planning and post-surgical treatment may not be currently possible due to the absence of knowledge correlating the model-predicted parameters to physiological variables. Links between these parameters to patient outcomes are yet to be established. This article reviews the recent application of computational modelling to understand the nasal structure-function relationship following surgery in patients with sinusitis and nasal obstruction.

Automated pancreas segmentation from computed tomography and magnetic resonance images: A systematic review.

The pancreas is a highly variable organ, the size, shape, and position of which are affected by age, sex, adiposity, the presence of diseases affecting the pancreas (e.g., diabetes, pancreatic cancer, pancreatitis) and other factors. Accurate automated segmentation of the pancreas has the potential to facilitate timely diagnosing and managing of diseases of the endocrine and exocrine pancreas. The aim was to systematically review studies reporting on automated pancreas segmentation algorithms derived from computed tomography (CT) or magnetic resonance (MR) images. The MEDLINE database and three patent databases were searched. Data on the performance of algorithms were meta-analysed, when possible. The algorithms were classified into one of four groups: multiorgan atlas-based, landmark-based, shape model-based, and neural network-based. A total of 13 cohorts suitable for meta-analysis were pooled to determine the performance of pancreas segmentation algorithms altogether using the Dice coefficient. These cohorts, comprising 1110 individuals, yielded a weighted mean Dice coefficient of 74.4%. Eight cohorts suitable for meta-analysis were pooled to determine the performance of pancreas segmentation algorithms altogether using the Jaccard index. These cohorts, comprising 636 individuals, yielded a weighted mean Jaccard index of 63.7%. Multiorgan atlas-based algorithms had a weighted mean Dice coefficient of 70.1% and a weighted mean Jaccard index of 59.8%. Neural network-based algorithms had a weighted mean Dice coefficient of 82.3% and a weighted mean Jaccard index of 70.1%. Studies using the other two types of algorithms were not meta-analysable. The above findings indicate that the automation of pancreas segmentation represents a considerable challenge as the performance of current automated pancreas segmentation algorithms is suboptimal. Adopting standardised reporting on performance of pancreas segmentation algorithms and encouraging the use of benchmark pancreas segmentation datasets will allow future algorithms to be tested and compared more easily and fairly.

Wave reflection in an anatomical model of the pulmonary circulation in local and global hypertensive disease.

Pulmonary hypertension is a disease of the pulmonary vasculature which can occur for many different reasons, including pathological remodeling of the pulmonary vessels and occlusion of these vessels (amongst others). Pulmonary hypertension can lead to right heart failure and significantly reduces the quality of life of patients living with the condition. It is difficult to distinguish clinically between different classifications of pulmonary hypertension, and doing so accurately is critical for the management of an individual's condition. In addition, different presentations of the disease (e.g. occlusion versus remodeling) can put different strains on the right heart, despite patients having very similar elevations in pulmonary artery pressure. In this study we use an anatomically based model of the pulmonary circulation to predict pressure and flow wave transmission and reflection in two different kinds of pulmonary hypertension - primary pulmonary hypertension, and chronic thrombotic pulmonary embolism (CTEPH), to enable analysis of the impact of disease type on impedance spectra in the main pulmonary artery.

Capturing complexity in pulmonary system modelling.

Respiratory disease is a significant problem worldwide, and it is a problem with increasing prevalence. Pathology in the upper airways and lung is very difficult to diagnose and treat, as response to disease is often heterogeneous across patients. Computational models have long been used to help understand respiratory function, and these models have evolved alongside increases in the resolution of medical imaging and increased capability of functional imaging, advances in biological knowledge, mathematical techniques and computational power. The benefits of increasingly complex and realistic geometric and biophysical models of the respiratory system are that they are able to capture heterogeneity in patient response to disease and predict emergent function across spatial scales from the delicate alveolar structures to the whole organ level. However, with increasing complexity, models become harder to solve and in some cases harder to validate, which can reduce their impact clinically. Here, we review the evolution of complexity in computational models of the respiratory system, including successes in translation of models into the clinical arena. We also highlight major challenges in modelling the respiratory system, while making use of the evolving functional data that are available for model parameterisation and testing.