Biomedical Computing

FEBIO - Finite Elements for Biomechanics and Biophysics

Jeff Weiss
Finite element analysis has become an indispensable tool for research and discovery in the biomedical sciences. Historically, the lack of an open software environment that was tailored to the needs of the field hampered research progress, dissemination of research and sharing of models and results. To address these issues, we developed the FEBio software suite, a FE framework designed specifically for analysis in biomechanics and biophysics, during our first funding period (2007-2011). FEBio employs mixture theory to account for the multi- constituent nature of biological tissues and fluids, unifying the classical fields of irreversible thermodynamics, solid mechanics, fluid mechanics, mass transport, chemical reactions and electrokinetics. During the second funding period (2012-2016), we implemented chemical reactions between constituents of a mixture and we broadened the target audience for FEBio by developing a plugin environment that made it easy to add features or interface other software with FEBio. During our third funding period (2016-2020), we developed a novel FE framework for simulation of compressible and incompressible CFD, extended this framework to enable analysis of FSI (Fluid-Structure Interaction) problems, and we enhanced algorithmic, analysis and numerical capabilities in FEBio by implementing efficient iterative linear solvers and preconditioners, new nonlinear solution strategies, and adaptive meshing. In this competing continuation application, we propose three aims: 1) Formulate and implement a computationally efficient damage and fatigue failure framework for fibrous tissues; 2) Extend our multiphysics algorithms to solute transport and reactions in fluid domains, and tissue growth and remodeling in fluid domains and at their interfaces with multiphasic domains; 3) Integrate the use of image data through our entire simulation pipeline, from model setup to model validation. These new capabilities will expand the applicability of FEBio to new fields of biomedical research, increasing our user base and facilitating scientific advancement.

Public Health Relevance
This project will greatly expand the simulation capabilities of the FEBio software suite, which enables computer simulations in the fields of biomechanics and biophysics. This software already supports over 10,000 scientists in their research. The outcomes of the proposed research will expand scientific discovery and innovation by biomedical simulation scientists in the areas of biomechanics and biophysics.

Regulation of Cell Adhesions by Mechanical Force

Tamara Bidone
This project aims to understanding how biological materials respond to mechanical force. Adhesion receptors are proteins in the plasma membrane that mediate cell binding to other cells or to the extracellular matrix. These receptors sense and transmit forces between cells and the external environment. Specifically, integrins transmit mechanical and biochemical information. This information critically guides life processes such as embryonic development, and also plays a role in diseases such as cancer, hypertension, osteoporosis and atherosclerosis. Integrins undergo drastic, long-range changes in response to biochemical signals or mechanical stresses. Exactly how integrin amino acid residues rearrange in response to force is largely unknown due to the technical challenges in observing shifts in residue positions and interactions at the single protein level. This hampers our ability to understand or modulate these processes that can lead to disease. This research will employ cell biology, computational simulations, biophysics, chemistry and material science approaches to elucidate the detailed pathways of force-dependent changes in integrins conformation and their roles in cell regulation.

The overarching objective of this research is to understand, at near-atomic level, how integrin conformation is modulated by mechanical force and how the resultant conformational transitions regulate cell behavior. The research team will develop new numerical tools to perform molecular dynamics simulations of integrins under force. These models will be used to predict mutations that stabilize or destabilize specific integrin states, which will then be tested experimentally for their effects on mechanosensing in live cells. By combining modeling with experiments, we will elucidate detailed pathways of force-dependent conformational transitions and their role in biology. This multidisciplinary approach will broaden the participation of diverse communities and underrepresented groups and positively impact engineering and material science education. Furthermore, this research may result in new approaches to design and produce biomimetic and smart materials for engineering applications.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Center for Integrative Biomedical Computing Legacy Transition

Chris Johnson, Rob MacLeod, Ross Whitaker
The Center for Integrative Biomedical Computing (CIBC) has an established, twenty year record as a Biomedical Technology Research Resource (BTRR) whose focus has been technical research and development in the area of image based modeling and simulation. The resulting legacy of scientific publications, novel algorithms and computational approaches and especially a suite of software tools and data sets can continue to support a vibrant research community even as the twilight requirements of the BTRR program dictate a time limit on NIH P41 funding. However, while already in heavy use, many of the existing, open source, software tools require additional hardening, documentation, and user-interface support to reach their full potential. More important, the software engineering and infrastructure, even though they represent the highest standards available from an academic center, are not yet robust and flexible enough to sustain easy maintenance and extension by a community outside the CIBC, a key element to ensuring the continued health and growth of the impact of these tools beyond the lifetime of the BTRR. Finally, there are existing collaborators and Driving Biomedical Project (DBP) partners whose research would suffer with the pending cessation of operations of the BTRR. We propose a transition plan to achieve long term sustainability of the software and data resources that the CIBC has produced. The goal of this plan is to complete the conversion of all the technical products achieved over the lifetime of the Center into well crafted, validated computer code with all the necessary support for both users and future maintainers of the code base. We seek to convert our open-source software into something far more valuable: community supported and sustained software. The conversion will require the assistance of a professional software house that is familiar with the necessary steps and with the technical domain of our center, Kitware Inc. Kitware is a company with a long history of assisting biomedical researchers to develop highly complex software packages and libraries, including VTK, ITK, 3D Slicer, and ParaView. Members of the CIBC have collaborated with Kitware in the past so there is a proven relationship on which to pursue the proposed transition plan. This combination of the domain specific knowledge of the specific tools within the twenty years of the CIBC, along with the ability of Kitware to support and deploy software and data, will ensure a successful transition and a lasting legacy from this BTRR.

Public Health Relevance
We propose a transition plan to achieve long term sustainability of the software and data resources that the Biomedical Technology Research Resource, Center for Integrative Biomedical Computing (CIBC), has produced. The goal of this plan is to complete, with the assistance of the software engineers at Kitware Inc, the conversion of all the technical products achieved over the lifetime of the CIBC into well crafted, validated computer code with all the necessary support for both users and future maintainers of the code base. We will combine domain specific knowledge of the specific tools within the twenty years of the CIBC, along with the ability of Kitware to support and deploy software and data, will ensure a successful transition and a lasting legacy from this BTRR.

Targeting Collagen Mechanical Damage Using Collagen Hybridizing Peptides

Jeff Weiss
Detection of Collagen Mechanical Damage using Collagen Hybridizing Peptides. Mechanical injury to load-bearing tissues leads to many clinically significant conditions (e.g. tendinosis, rotator cuff disease) but we have limited understanding of the injury process of tissues that are damaged by mechanical stress. The overall goal of the proposed research is to gain new understanding of the biomechanics of load bearing collagenous tissues by developing the collagen hybridizing peptide (CHP) technology into a new mechanical damage detection method. CHP has been reported to bind to denatured collagen strands originating from protease activity or by mechanical damage in a manner similar to primer binding to melted DNA during PCR. We propose to substantially expand the capabilities of CHP damage detection by developing new CHPs that are smaller for faster diffusion into dense musculoskeletal tissues and exhibit accelerated binding kinetics to allow faster damage reporting. We will also develop a new CHP that only fluoresces upon binding with collagen, eliminating the need to stain and wash tissues and enabling the CHP to serve as a damage gauge in overloaded tissues. We will then develop optimized protocols for the use of the existing and new CHPs, determine the relationship between collagen fibril strain and CHP binding in musculoskeletal soft tissues, and quantitatively compare CHP targeting to other techniques. Finally, we will apply CHP targeting to elucidate the relationship between tissue level mechanical loading and mechanical damage to collagen at the molecular level. We will focus on two important musculoskeletal tissues of considerable clinical relevance: tendons and articular cartilage. Considering the wide-spread impact of collagen damage in musculoskeletal injuries and diseases, in- depth understanding of the relationships between molecular level collagen damage and mechanical overloading will provide new insights into the biomechanics of load-bearing tissues as well as help develop new diagnostics and therapies for managing musculoskeletal disorders.

Public Health Relevance
Detection of Collagen Mechanical Damage using Collagen Hybridizing Peptides. The proposed work will develop new and improved peptide-based probes that can hybridize to mechanically damaged collagen with high efficiency; these probes will be used to elucidate the mechanisms of stress-induced injury in tendon and articular cartilage. Fibrous collagen is the major component of almost every load-bearing tissue, and molecular-level collagen damage is associated with permanent mechanical damage in those tissues. The outcome of this research will help manage common musculoskeletal injuries and disorders by deepening our understanding of mechanical damage in tissues and by identifying opportunities for new diagnostics and therapies.

Neovessel Guidance in Angiogenesis

Jeff Weiss
Neovessel guidance in angiogenesis Vascular connectivity between adjacent vessel beds within and between tissue compartments is an essential aspect of any successful neovascularization process. To establish new connections, growing neovessels must locate other vascular elements during angiogenesis, often crossing matrix and other tissue-associated boundaries and interfaces. This is perhaps best highlighted in situations of tissue grafting (e.g. tissue free flaps) during which the vasculature of the graft must cross the graft-host tissue interface before connecting to the surrounding host circulation in order for the graft to survive. An inability for a tissue graft vasculature to connect to the host vasculature is the primary cause for tissue graft failure and necrosis. How growing neovessels traverse any tissue interface, whether part of the native tissue structure or secondary to a grafting procedure, is not known. In a series of preliminary experiments, we have determined that actively growing neovessels are unable to spontaneously cross a stroma-stroma interface during angiogenesis. Our published findings that tissue stromal biomechanics, specifically the biophysical aspects of the matrix, has a profound influence on the direction and branching of growing neovessels suggests that this interface is biomechanically incompatible with interface invasion. In addition, we have evidence that a sub-population of tissue-resident, stromal macrophages facilitates neovessel invasion of the stromal interface during angiogenesis through a VEGF-A-dependent process. Importantly, it appears that stromal cells need to migrate across the interface in order to promote interface invasion by the neovessels, suggesting that spatial gradients of VEGF-A are required. Based on these observations, we hypothesize that the graded angiogenic factor signals overcome the biomechanical barriers to directed angiogenesis caused by a stroma-stroma interface to promote interface neovascular invasion. The project will involve a combination of in vitro and computational models of angiogenesis across tissue-tissue boundaries to test this hypothesis and determine the mechanism by which stromal cells use VEGF-A signaling to regulate neovessel behavior in coordination with stroma biomechanical dynamics. These studies will provide new insights into a poorly understood aspect of vascular biology and tissue-vascular dynamics as well as create opportunities for therapeutic strategies to facilitate tissue healing, improving angiogenesis-based treatments, and tissue grafting/transplantation.

Neovessel guidance in angiogenesis Tissue repair and grafting is often compromised by poor vascular connectivity between adjacent vascular beds. This project will investigate the regulation of angiogenic neovessel invasion into adjacent tissues, a necessary aspect of vascular connectivity, by resident stromal cells. By defining the mechanisms underlying facilitation of vessel connectivity by resident stromal cells in combination with inherent tissue biomechanical dynamics, we will provide insight into novel therapeutic strategies for tissue repair and grafting/transplantation.

Multi-Tensor Decompositions for Personalized Cancer Diagnostics and Prognostics

Orly Alter
Recurring DNA copy-number alterations (CNAs) have been recognized as a hallmark of cancer for >100 years, yet what these alterations imply about a tumor's pathogenesis and a patient's diagnosis, prognosis, and treatment remains poorly understood. This is despite the growing number of large-scale multidimensional datasets recording different aspects of a single disease, e.g., in the Cancer Genome Atlas (TCGA), and due to a fundamental need for mathematical frameworks that can create one coherent model from such multiple datasets arranged in multiple tensors of matched columns, e.g., patients, platforms, and tissues, but independent rows, e.g., probes. For example, our recent comparative modeling (by using a data-driven two-matrix spectral decomposition) of patient-matched glioblastoma (GBM) brain tumor and normal blood genomic profiles from TCGA (arranged in two matrices, of matched columns but independent rows) uncovered a previously unknown global pattern of tumor-exclusive CNAs that is correlated with, and possibly causally related to, GBM survival and response to chemotherapy. The data had been publicly available since 2008, but this signature remained unknown until we applied our comparative modeling in 2012. Survival analyses showed, and computationally validated, that the signature performs better than, and is statistically independent of, age, the best indicator of GBM survival for >50 years, and existing GBM pathology laboratory tests. A new test for GBM based upon this signature is pending an experimental re-validation at the Associated Regional and University Pathologists (ARUP) Laboratories, Inc., a nonprofit reference laboratory of the Department of Pathology at the University of Utah. In this NCI U01 project, our multidisciplinary team of researchers from the Departments of Bioengineering, Mathematics, and Pathology, the Scientific Computing and Imaging (SCI) Institute, and the Huntsman Cancer Institute (HCI) at the University of Utah, aims to (i) define, and study the properties of data- driven multi-tensor spectral decompositions; (ii) use these to model patient-, platform-, and tissue-matched but probe-independent TCGA genomic profiles, and gain biological and medical insights into the genotype- phenotype relations in lower-grade astrocytoma (LGA) brain cancer, ovarian serous cystadenocarcinoma (OV), and lung squamous cell carcinoma; and (iii) enable translation of these insights into pathology laboratory tests, by experimentally testing the computational predictions of the existing GBM model, as well as the novel LGA and OV models by using Utah samples. Ultimately, this project will bring physicians a step closer to one day being able to predict and control the progression of cell division and cancer as readily as NASA engineers plot the trajectories of spacecraft today.

Public Health Relevance
Recurring DNA copy-number alterations have been recognized as a hallmark of cancer for >100 years, yet what these alterations imply about a tumor's pathogenesis and a patient's diagnosis, prognosis, and treatment remains poorly understood. In this NCI U01 project, we will (i) develop generalizations of the mathematical frameworks that underlie the theoretical description of the physical world, (ii) use these frameworks to model patient-matched datasets from the Cancer Genome Atlas, and gain biological and medical insights into the genotype-phenotype relations in cancer, and (iii) translate these insights into pathology laboratory tests. Ultimately this projct will bring physicians a step closer to one day being able to predict and control the progression of cell division and cancer as readily as NASA engineers plot the trajectories of spacecraft today.

New 3d and 4d Image-Based Clinical Measures of Atrial Structural and Functional Remodeling in Atrial Fibrillation

Joshua Cates, Shireen Youssef Elhabian
This proposal focuses on the development of new clinical measures of atrial shape for better individualized patient care of atrial fibrillation and advances the state-of-the art in computational methods for the statistical analysis of shape in clinical image data. Atrial fibrillation (AF) is the most common cardiac arrhythmia in adults, with over 400,000 U.S. hospital admissions per year, annual costs in the U.S. of $6-7 billion, and an almost two-fold increase in the risk of mortality. We will develop and test new cardiac shape-based predictors of AF outcomes in a large patient population, including patients enrolled in the fully-funded multicenter DECAAF II clinical trial. This proposal will impact clinical science and standards of care for AF patients through (1) the development of a new MRI-based measure of dynamic cardiac shape and associated deformations over the cardiac cycle and investigation of the dynamic (4D) structural changes to the atria after catheter ablation; (2) the development of shape-based predictors of AF treatment outcomes and risk factors for stroke; and (3) shape analysis of the atria and prospective evaluation of shape-based predictive indices in the DECAAF II clinical trial population. Technological contributions of this work will impact the field of biomedical shape analysis through the development of methods for analysis of spatiotemporal (4D) anatomical shape and the development of more powerful modeling algorithms for the highly variable anatomy of the heart. This project combines the clinical experience and research infrastructure of the University of Utah's Comprehensive Arrhythmia Research and Management (CARMA) Center with the computational resources and expertise of the Scientific Computing and Imaging (SCI) Institute and leverages our unique access to a large cohort of MRI image data from AF patients enrolled in the multicenter DECAAF II clinical trial. Our team consists of an accomplished mix of computer scientists, electrophysiologists, and MRI physicists.

Public Health Relevance
The goals of this proposal are to develop non-invasive image-based measures of atrial shape (geometry) and function that are clinically useful in planning patient treatment and managing risk of adverse outcomes such as stroke. Atrial fibrillation affects many millions of people in the United States, incurs annual costs of $6-7 billion, and is associated with significant mortality and morbidity.

Lab-to-Lab Training and Dissemination for the Febio Software Suite

Jeff Weiss
Finite element analysis has become an indispensable tool for research and discovery in the biomedical sciences. Historically, the lack of a software environment that was tailored to the needs of the field hampered research progress, dissemination of research and sharing of models and results. Investigators previously relied on commercial software, but these packages were not geared toward biological applications, were difficult to verify, and precluded the easy addition, extension and sharing of new features. To address these issues, we developed the FEBio software suite, a FE framework designed specifically for analysis in biomechanics and biophysics. FEBio employs mixture theory to account for the multi-constituent nature of soft and hard tissues, unifying the classical fields of irreversible thermodynamics, solid mechanics, fluid mechanics, mass transport, chemical reactions and electrokinetics. Despite the popularity of the FEBio software suite, we have never been able to fully address the needs of our user base in terms of dissemination and training. The overall aim of this project is to develop an integrated software environment for FEBio to deliver new approaches for digital training and dissemination to our user community. This will be accomplished through the following specific aims: 1) Create a unified graphical software environment that encompasses the entire FEBio analysis pipeline. 2) Develop a model sharing environment for FEbio. 3) Develop digital training materials and virtual training courses to serve and expand the FEBio community. We will partner with laboratories that already use FEBio to evaluate these advancements in training and dissemination. These new mechanisms for training and dissemination will improve the accessibility of FEBio's broad simulation capabilities, creating a larger, more collaborative user group that spans a greater range of technical disciplines.

Public Health Relevance
This project will develop lab-to-user training and dissemination tools for the FEBio software suite. This software supports nearly 10,000 scientists in the fields of biomechanics, biophysics and biomedical engineering in general. The materials developed under this project will support the training of the next generation of biomedical simulation scientists in the areas of biomechanics and biophysics.

Novel Characterization and Detection Techniques for Acute Myocardial Ischemia

Brian Zenger
The most common reason for a patient to visit the emergency department is chest pain caused by myocardial ischemia. Myocardial ischemia develops from inadequate perfusion of myocardial tissue and indicates the presence of surprising range of conditions, including coronary artery disease, coronary microvascular dysfunction, Takotsubo cardiomyopathy, and coronary artery dissection, and other potentially fatal cardiac diseases. Current noninvasive tests to detect ischemia may be limited by an incomplete understanding of the mecha- nisms used to induce the stress (exercise or pharmacological agents) or the limited diagnostic data acquired to visualize ischemia (12-lead electrocardiogram or ultrasound). These limitations increase the risk for failed detection of deadly cardiac diseases and explain the unsatisfactory accuracy of current diagnostic and monitor- ing approaches. The goal of this project is to leverage recent experimental findings and breakthroughs about the onset and progression of ischemia as the basis for a comprehensive re-evaluation of acute myocardial ischemia. We will develop and use experimental models and computer simulations to devise better techniques to detect acute myocardial ischemia and localize it within the ventricles. We will build on our experience with our novel large-animal experimental models of ischemia and electrocardiographic imaging (ECGI) techniques to measure and characterize the electrical changes that arise during acute myocardial ischemia created from various types of ischemic stress. We will use these measurements to characterize different types of cardiac stress and implement novel ECGI techniques to detect and localize ischemic sources within the heart.

Public Health Relevance
Myocardial ischemia arises under conditions of inadequate blood supply to the heart and is associated with a surprisingly broad range of conditions, such as coronary artery disease, ischemic cardiomyopathy, or coronary artery dissection. The most common symptom (chest pain) is also the most frequent reason for visits to the emergency department of hospitals. The goal of this project is to create and validate, through controlled experimental preparations and computer modeling approaches, a painless, inexpensive, and noninvasive detection system based on the electrocardiogram to localize regions of myocardial ischemia within the heart.

Integration of Uncertainty Quantification with SCIRun Bioelectric Field Simulation Pipeline

Rob MacLeod
This project harnesses mature and modern techniques for uncertainty quantification (UQ) from the computational science and mathematics communities and incorporates them into the established biomedical simulation software framework SCIRun. Just as bioelectric simulation models have become increasingly accurate at representing physical processes and geometry, so also has it become increasingly important to understand the uncertainty stemming from ignorance about precise values of model parameters. To our knowledge, success of this project represents one of the first developments of folding in modern UQ techniques into widely used, freely available software pipelines for simulation of bioelectric fields in heart and brain applications. Successful incorporation of established and proven UQ tools into SCIRun will substantially increase the utility of simulation based biomedical predictions in medical practice. Although there is a need in the research community for such tools, the number and nature of users makes commercial development very unlikely. There are no commercial tools specifically for bioelectric field modeling and while some industry uses these tools, their high intellectual cost and limited market keep them well below the threshold for commercial software.

Public Health Relevance
This project harnesses mature and modern techniques that can quantify statistical uncertainty in biomedical simulations by integrating these techniques into the well-established biomedical modeling software framework, SCIRun. Biomedical simulation has become increasingly accurate at representing functions of the body, both in health and disease, however such simulations rarely include any estimates of the uncertainty stemming from ignorance about precise values of model parameters. Our proposal represents one of the first to integrate modern methods capable of quantifying such uncertainty into widely used, freely available software pipelines for simulation of bioelectric fields from the heart and brain.