Mechanical Ventilator Parameter Estimation for Lung Health through Machine Learning.

Patients whose lungs are compromised due to various respiratory health concerns require mechanical ventilation for support in breathing. Different mechanical ventilation settings are selected depending on the patient's lung condition, and the selection of these parameters depends on the observed patient response and experience of the clinicians involved. To support this decision-making process for clinicians, good prediction models are always beneficial in improving the setting accuracy, reducing treatment error, and quickly weaning patients off the ventilation support. In this study, we developed a machine learning model for estimation of the mechanical ventilation parameters for lung health. The model is based on inverse mapping of artificial neural networks with the Graded Particle Swarm Optimizer. In this new variant, we introduced grouping and hierarchy in the swarm in addition to the general rules of particle swarm optimization to further improve its prediction performance of the mechanical ventilation parameters. The machine learning model was trained and tested using clinical data from canine and feline patients at the University of Georgia College of Veterinary Medicine. Our model successfully generated a range of parameter values for the mechanical ventilation applied on test data, with the average prediction values over multiple trials close to the target values. Overall, the developed machine learning model should be able to predict the mechanical ventilation settings for various respiratory conditions for patient's survival once the relevant data are available.

Identification of predictive MRI and functional biomarkers in a pediatric piglet traumatic brain injury model.

Traumatic brain injury (TBI) at a young age can lead to the development of long-term functional impairments. Severity of injury is well demonstrated to have a strong influence on the extent of functional impairments; however, identification of specific magnetic resonance imaging (MRI) biomarkers that are most reflective of injury severity and functional prognosis remain elusive. Therefore, the objective of this study was to utilize advanced statistical approaches to identify clinically relevant MRI biomarkers and predict functional outcomes using MRI metrics in a translational large animal piglet TBI model. TBI was induced via controlled cortical impact and multiparametric MRI was performed at 24 hours and 12 weeks post-TBI using T1-weighted, T2-weighted, T2-weighted fluid attenuated inversion recovery, diffusion-weighted imaging, and diffusion tensor imaging. Changes in spatiotemporal gait parameters were also assessed using an automated gait mat at 24 hours and 12 weeks post-TBI. Principal component analysis was performed to determine the MRI metrics and spatiotemporal gait parameters that explain the largest sources of variation within the datasets. We found that linear combinations of lesion size and midline shift acquired using T2-weighted imaging explained most of the variability of the data at both 24 hours and 12 weeks post-TBI. In addition, linear combinations of velocity, cadence, and stride length were found to explain most of the gait data variability at 24 hours and 12 weeks post-TBI. Linear regression analysis was performed to determine if MRI metrics are predictive of changes in gait. We found that both lesion size and midline shift are significantly correlated with decreases in stride and step length. These results from this study provide an important first step at identifying relevant MRI and functional biomarkers that are predictive of functional outcomes in a clinically relevant piglet TBI model. This study was approved by the University of Georgia Institutional Animal Care and Use Committee (AUP: A2015 11-001) on December 22, 2015.

Physiochemical Effects of Nanoparticles on Cell Nuclear Complex Pore Transport: A Coarse-Grained Computational Model.

Understanding and controlling the interaction between nanoparticles and cell nuclei is critical to the development of the biomedical applications such as gene delivery, cellular imaging, and tumor therapy. Recent years have witnessed growing evidence that the size, shape, and the grafting density of the karyopherins ligands of nanoparticles provide a significant influence on the uptake mechanism of nanoparticles into cells; however, there is a lack of investigation into how these physical factors play a role in cellular nuclear uptake and how the nanoparticle enters the nucleus. Here, we build a computational framework to parametrically evaluate the effects of the size, shape, and the grafting density of the karyopherins ligands of designed nanoparticles on their transport through the nuclear pore complex of a cell nucleus so as to provide a novel scheme for nanoparticle design and precise nucleus-targeted therapy. Simulation results indicate that smaller spherical nanoparticles need to overcome a lower energy barrier than larger ones and also that nanoparticles with large grafting density exhibited greatly altered dynamics during the active transport process. Moreover, we observed that the shape and morphology of nanoparticles unambiguously determined their nuclear uptake pathways. Nuclear uptake is determined by an intricate interplay between physicochemical particle properties and nucleus properties. Our work provides a systematic understanding for nuclear uptake of nanoparticles, viruses, and bacteria and opens up a controllable design strategy for manipulating nanoparticle-nucleus interaction, with numerous applications in medicine, bioimaging, and biosensing.

Aging Effects on Alveolar Sacs Under Mechanical Ventilation.

Alveolar sacs are primarily responsible for gas exchange in the human respiratory system and lose their functionality with aging. Three-dimensional (3D) models of young and old human alveolar sacs were constructed and fluid-solid interaction was employed to investigate the contribution of age-related changes to decline in alveolar sacs function under mechanical ventilation (MV). Simulation results illustrated that compliance and pressure reduced in the alveolar sacs of the elderly adults, and they have to work harder to breathe. Morphological changes were found to be mainly responsible for the decline in alveolar sacs function. Influence of individual differences on the alveolar sacs function was negligible and 95% confidence intervals for compliance and work of breathing (WOB) using measures from different individuals also support this finding. Moreover, higher mortality risk was recorded for elderly adults who undergo MV. Specifically, ventilator devices setting has been identified as a potential parameter for compromising respiratory function in the elderly adults. Volume-controlled ventilation applied less pressure, whereas, pressure-controlled ventilation resulted in higher compliance in the alveolar sacs and decreased WOB. Sensitivity of alveolar sacs to ventilator setting under the volume-controlled mode illustrated that increasing breathing frequency and decreasing the ratio of inhalation to exhalation times and TV caused an increase in alveolar sacs expansion and compliance in older patients. Results from this study can help clinicians to develop individualized and effective ventilator protocols and to improve respiratory function in the elderly adults.

An integrated parametric model for MT self-assembly formation analysis.

Self-assembly is a ubiquitous, naturally occurring, robust process in many living organisms. Microtubule (MT), a self-organization system assemble itself into functional units by attaching to cellular structures. Modeling microtubule self-organization is of interest as microtubule forms a network of protein filaments that is critical to many processes in eukaryotic cells. In this paper, we propose an optimization framework that considers MT self-assembly starting from alpha (α) and beta (ß) tubulins as basic building blocks in the self-organization of MT. Using this framework we present separate analysis of MT self-assembly strength by considering two aspects of MT self-assembly. First, the affinity factor distribution between neighboring tubulins of an MT is considered for the analysis. Second, this paper also present an analysis of structural stability considering geometric parameter distribution of tubulins within an MT. We present separate algorithms for the analysis in detail. The proposed models show convergence and robustness under random initialization and thus justify the effectiveness of the proposed convergence criteria for stability analysis of MT self-organization. The proposed algorithms show the ability to emulate MT self-assembly from random initial configurations.

Evaluation of Ventilation-Induced Lung Inflammation Through Multi-Scale Simulations.

Ventilation-induced lung injury is a common problem faced by patients with respiratory problems who require mechanical ventilation (MV). This injury may lead to a greater chance of developing or exacerbating the acute respiratory distress syndrome which further complicates the therapeutic use of MV. The chain of events begins with the MV initiating an immune response that leads to inflammation induced tissue material alteration (stiffening) and eventually the loss of lung resistance. It is clear from this sequence of events that the phenomenon of ventilation induced injury is multi-scale by nature and, hence, requires holistic analysis involving simulations and informatics. An effective approach to this problem is to break it down into several major physical models. Each physical model is developed separately and can be seen as a component in a larger system that comprises the scale of the problem being investigated. In this paper, a multi-scale system consisting of breathing mechanics, tissue deformation, and cellular mechanics models is developed to assess the immune response. To demonstrate the potential of the model, a fluid-solid model is employed for breathing mechanics, a plane-strain elasticity model is applied to assess tissue deformation, and a cellular automata (CA) model is developed to account for immune response. A case study of three lower airways is presented. The CA model shows that this increased the immune response by five times, which correlates with alteration in the tissue microstructure. This alteration in turn is reflected in the material constant value obtained in the tissue mechanics model. However, the changes in strain rates in the airways after inflammation (and hence, lung compliance) were not as significant as the rates of change in immune response. Finally, results from the fluid-solid model demonstrate its potential for airflow characterization caused by tissue deformation that could lead to disease identification.

Simulation of Healing Threshold in Strain-Induced Inflammation Through a Discrete Informatics Model.

Respiratory diseases such as asthma and acute respiratory distress syndrome as well as acute lung injury involve inflammation at the cellular level. The inflammation process is very complex and is characterized by the emergence of cytokines along with other changes in cellular processes. Due to the complexity of the various constituents that makes up the inflammation dynamics, it is necessary to develop models that can complement experiments to fully understand inflammatory diseases. In this study, we developed a discrete informatics model based on cellular automata (CA) approach to investigate the influence of elastic field (stretch/strain) on the dynamics of inflammation and account for probabilistic adaptation based on statistical interpretation of existing experimental data. Our simulation model investigated the effects of low, medium, and high strain conditions on inflammation dynamics. Results suggest that the model is able to indicate the threshold of innate healing of tissue as a response to strain experienced by the tissue. When strain is under the threshold, the tissue is still capable of adapting its structure to heal the damaged part. However, there exists a strain threshold where healing capability breaks down. The results obtained demonstrate that the developed discrete informatics based CA model is capable of modeling and giving insights into inflammation dynamics parameters under various mechanical strain/stretch environments.

Quantification of Age-Related Lung Tissue Mechanics under Mechanical Ventilation.

Elderly patients with obstructive lung diseases often receive mechanical ventilation to support their breathing and restore respiratory function. However, mechanical ventilation is known to increase the severity of ventilator-induced lung injury (VILI) in the elderly. Therefore, it is important to investigate the effects of aging to better understand the lung tissue mechanics to estimate the severity of ventilator-induced lung injuries. Two age-related geometric models involving human bronchioles from generation G10 to G23 and alveolar sacs were developed. The first is for a 50-year-old (normal) and second is for an 80-year old (aged) model. Lung tissue mechanics of normal and aged models were investigated under mechanical ventilation through computational simulations. Results obtained indicated that lung tissue strains during inhalation (t = 0.2 s) decreased by about 40% in the alveolar sac (G23) and 27% in the bronchiole (G20), respectively, for the 80-year-old as compared to the 50-year-old. The respiratory mechanics parameters (work of breathing per unit volume and maximum tissue strain) over G20 and G23 for the 80-year-old decreased by about 64% (three-fold) and 80% (four-fold), respectively, during the mechanical ventilation breathing cycle. However, there was a significant increase (by about threefold) in lung compliance for the 80-year-old in comparison to the 50-year-old. These findings from the computational simulations demonstrated that lung mechanical characteristics are significantly compromised in aging tissues, and these effects were quantified in this study.

Bio-Inspired Multi-Functional Drug Transport Design Concept and Simulations.

In this study, we developed a microdevice concept for drug/fluidic transport taking an inspiration from supramolecular motor found in biological cells. Specifically, idealized multi-functional design geometry (nozzle/diffuser/nozzle) was developed for (i) fluidic/particle transport; (ii) particle separation; and (iii) droplet generation. Several design simulations were conducted to demonstrate the working principles of the multi-functional device. The design simulations illustrate that the proposed design concept is feasible for multi-functionality. However, further experimentation and optimization studies are needed to fully evaluate the multifunctional device concept for multiple applications.

Aging effects on airflow dynamics and lung function in human bronchioles.

BACKGROUND AND OBJECTIVE: The mortality rate for patients requiring mechanical ventilation is about 35% and this rate increases to about 53% for the elderly. In general, with increasing age, the dynamic lung function and respiratory mechanics are compromised, and several experiments are being conducted to estimate these changes and understand the underlying mechanisms to better treat elderly patients. MATERIALS AND METHODS: Human tracheobronchial (G1 ~ G9), bronchioles (G10 ~ G22) and alveolar sacs (G23) geometric models were developed based on reported anatomical dimensions for a 50 and an 80-year-old subject. The aged model was developed by altering the geometry and material properties of the model developed for the 50-year-old. Computational simulations using coupled fluid-solid analysis were performed for geometric models of bronchioles and alveolar sacs under mechanical ventilation to estimate the airflow and lung function characteristics. FINDINGS: The airway mechanical characteristics decreased with aging, specifically a 38% pressure drop was observed for the 80-year-old as compared to the 50-year-old. The shear stress on airway walls increased with aging and the highest shear stress was observed in the 80-year-old during inhalation. A 50% increase in peak strain was observed for the 80-year-old as compared to the 50-year-old during exhalation. The simulation results indicate that there is a 41% increase in lung compliance and a 35%-50% change in airway mechanical characteristics for the 80-year-old in comparison to the 50-year-old. Overall, the airway mechanical characteristics as well as lung function are compromised due to aging. CONCLUSION: Our study demonstrates and quantifies the effects of aging on the airflow dynamics and lung capacity. These changes in the aging lung are important considerations for mechanical ventilation parameters in elderly patients. Realistic geometry and material properties need to be included in the computational models in future studies.

Investigation of Diffusion Characteristics through Microfluidic Channels for Passive Drug Delivery Applications.

Microfluidics has many drug delivery applications due to the ability to easily create complex device designs with feature sizes reaching down to the 10s of microns. In this work, three different microchannel designs for an implantable device are investigated for treatment of ocular diseases such as glaucoma, age-related macular degeneration (AMD), and diabetic retinopathy. Devices were fabricated using polydimethylsiloxane (PDMS) and soft lithography techniques, where surface chemistry of the channels was altered using 2-[methoxy(polyethyleneoxy)propyl]trimethoxysilane (PEG-silane). An estimated delivery rate for a number of common drugs was approximated for each device through the ratio of the diffusion coefficients for the dye and the respective drug. The delivery rate of the model drugs was maintained at a physiological condition and the effects of channel design and surface chemistry on the delivery rate of the model drugs were recorded over a two-week period. Results showed that the surface chemistry of the device had no significant effect on the delivery rate of the model drugs. All designs were successful in delivering a constant daily dose for each model drug.

Transient Mechanical Response of Lung Airway Tissue during Mechanical Ventilation.

Patients with acute lung injury, airway and other pulmonary diseases often require Mechanical Ventilation (MV). Knowledge of the stress/strain environment in lung airway tissues is very important in order to avoid lung injuries for patients undergoing MV. Airway tissue strains responsible for stressing the lung's fiber network and rupturing the lung due to compliant airways are very difficult to measure experimentally. Multi-level modeling is adopted to investigate the transient mechanical response of the tissue under MV. First, airflow through a lung airway bifurcation (Generation 4-6) is modeled using Computational Fluid Dynamics (CFD) to obtain air pressure during 2 seconds of MV breathing. Next, the transient air pressure was used in structural analysis to obtain mechanical strain experienced by the airway tissue wall. Structural analysis showed that airway tissue from Generation 5 in one bifurcation can stretch eight times that of airway tissue of the same generation number but with different bifurcation. The results suggest sensitivity of load to geometrical features. Furthermore, the results of strain levels obtained from the tissue analysis are very important because these strains at the cellular-level can create inflammatory responses, thus damaging the airway tissues.

Effect of mechanical ventilation waveforms on airway wall shear.

Better understanding of airway wall shear stress/strain rate is very important in order to prevent inflammation in patients undergoing mechanical ventilation due to respiratory problems in intensive-care medicine. The objective of this study was to investigate the role of mechanical ventilation waveforms on airway wall shear/strain rate using computational fluid dynamics analysis. Six different waveforms were considered to investigate the airway wall shear stress (WSS) from fluid dynamics analysis for the airway geometry of two-to-three generations. The simulation results showed that Original with Sine Inhale Waveform (OSIW) produced the highest WSS value and the Near True Sine Waveform produced the lowest WSS value. Also, the Original with Sine Inhale Waveform and the Short Sine Inhale with Long Sine Exhale Waveform (SSILSEW) produced a higher shear strain rate in comparison to the Original Waveform (OW). These results, combined with optimization, suggest that it is possible to develop a set of mechanical ventilation waveform strategies to avoid inflammation in the lung.

A biomechanical model of wound contraction and scar formation.

We propose a biomechanical model for investigating wound contraction mechanism and resulting scarring. Extracellular matrix is modeled as fiber-reinforced anisotropic soft tissue, with its elastic properties dynamically changing with the density and orientation of collagen fibers. Collagen fibers are deposited by fibroblasts infiltrating the wound space, and are dynamically aligned with both migrating fibroblasts and tissue residing tension field. Our new 2D hybrid agent-based model provides a comprehensive platform for examining the mechanobiology in wound contraction and scar formation. Simulation results are consistent with experimental observations and are able to reveal the effects of wound morphology and mechanical environment on contraction patterns. Our model results support the hypothesis that scar formation is the product of collagen fiber synthesis and alignment in the presence of the tensile stress field generated by a wound contraction process.

Analysis for stress environment in the alveolar sac model.

Better understanding of alveolar mechanics is very important in order to avoid lung injuries for patients undergoing mechanical ventilation for treatment of respiratory problems. The objective of this study was to investigate the alveolar mechanics for two different alveolar sac models, one based on actual geometry and the other an idealized spherical geometry using coupled fluid-solid computational analysis. Both the models were analyzed through coupled fluid-solid analysis to estimate the parameters such as pressures/velocities and displacements/stresses under mechanical ventilation conditions. The results obtained from the fluid analysis indicate that both the alveolar geometries give similar results for pressures and velocities. However, the results obtained from coupled fluid-solid analysis indicate that the actual alveolar geometry results in smaller displacements in comparison to a spherical alveolar model. This trend is also true for stress/strain between the two models. The results presented indicate that alveolar geometry greatly affects the pressure/velocities as well as displacements and stresses/strains.

Cellular automata modeling of pulmonary inflammation.

Better understanding of the acute/chronic inflammation in airways is very important in order to avoid lung injuries for patients undergoing mechanical ventilation for treatment of respiratory problems. Local lung inflammation is triggered by many mechanisms within the lung, including pathogens. In this study, a cellular automata based model (CA) for pulmonary inflammation that incorporates biophysical processes during inflammatory responses was developed. The developed CA results in three possible outcomes related to homeostasis (healing), persistent infection, and resolved infection with high inflammation (inflamed state). The results from the model are validated qualitatively against other existing computational models. A sensitivity analysis was conducted on the model parameters and the outcomes were assessed. Overall, the model results showed possible outcomes that have been seen in clinical practice and animal models. The present model can be extended to include inflammation resulting from damage tissue and eventually to model inflammation resulting from acute lung injury and multiple organ dysfunction syndromes in critical illness and injury.

Design of an implantable device for ocular drug delivery.

Ocular diseases, such as, glaucoma, age-related macular degeneration (AMD), diabetic retinopathy, and retinitis pigmentosa require drug management in order to prevent blindness and affecting million of adults in USA and worldwide. There is an increasing need to develop devices for drug delivery to address ocular diseases. This study focuses on the design, simulation, and development of an implantable ocular drug delivery device consisting of micro-/nanochannels embedded between top and bottom covers with a drug reservoir made from polydimethylsiloxane (PDMS) which is silicon-based organic and biodegradable polymer. Several simulations were carried out with six different micro-channel configurations in order to see the feasibility for ocular drug delivery applications. Based on the results obtained, channel design of osmotic I and osmotic II satisfied the diffusion rates required for ocular drug delivery. Finally, a prototype illustrating the three components of the drug delivery design is presented. In the future, the device will be tested for its functionality and diffusion characteristics.

Simulation of the effect of airway disease on respiratory airways.

Airway disease such as tumours and asthma lead to lung injuries. Therefore, a better understanding of airway mechanics parameters is very important to avoid lung injuries in patients undergoing mechanical ventilation for treatment of respiratory problems in intensive-care medicine as well as pulmonary medicine. The objective of this study was to investigate the role of airway diseases such as asthma and tumours on airway mechanics parameters using coupled fluid-solid computational analysis. The results obtained indicate that both tumours and asthma greatly affect the airway mechanics parameters (airflow velocity increased by about 15% and the strains increased by about 40%). Strain results of this study highlight significant changes in levels of airway parameters, which may translate into higher health risk associated with airway tumours and the asthmatic airways. These results combined with optimization suggest that it is possible to develop mechanical ventilation protocols to avoid lung injuries in patients.

Assessment of mechanical ventilation parameters on respiratory mechanics.

Better understanding of airway mechanics is very important in order to avoid lung injuries for patients undergoing mechanical ventilation for treatment of respiratory problems in intensive-care medicine, as well as pulmonary medicine. Mechanical ventilation depends on several parameters, all of which affect the patient outcome. As there are no systematic numerical investigations of the role of mechanical ventilation parameters on airway mechanics, the objective of this study was to investigate the role of mechanical ventilation parameters on airway mechanics using coupled fluid-solid computational analysis. For the airway geometry of 3 to 5 generations considered, the simulation results showed that airflow velocity increased with increasing airflow rate. Airway pressure increased with increasing airflow rate, tidal volume and positive end-expiratory pressure (PEEP). Airway displacement and airway strains increased with increasing airflow rate, tidal volume and PEEP form mechanical ventilation. Among various waveforms considered, sine waveform provided the highest airflow velocity and airway pressure while descending waveform provided the lowest airway pressure, airway displacement and airway strains. These results combined with optimization suggest that it is possible to obtain a set of mechanical ventilation strategies to avoid lung injuries in patients.

Tissue strains induced in airways due to mechanical ventilation.

Better understanding of the stress/strain environment in airway tissues is very important in order to avoid lung injuries for patients undergoing mechanical ventilation for treatment of respiratory problems. Airway tissue strains responsible for stressing the lung's fiber network and rupturing the lung due to compliant airways are very difficult to measure experimentally. A computational model that incorporates the heterogeneity of the airways was developed to study the effects of airway tissue material properties on strain distributions within each layer of the airway wall. The geometry and boundary conditions of the tissue strain analysis were obtained from the organ-level analysis model. Two sets of airway tissue properties (heterogeneous and homogeneous) were considered in order to estimate the strain levels induced within the tissue. The simulation results showed that the homogeneous model overestimated the maximum strain in the mucosa layer and underestimated the maximum strain in the smooth muscle and cartilage layers. The results of strain levels obtained from the tissue analysis are very important because these strains at the cellular-level can create inflammatory responses, thus damaging the airway tissues.