TRD - Sim and Est

点击翻译按钮获取中文摘要

3.1 Summary/Abstract In this technical core of the Center we will study both brain and cardiac bioelectric activity and, for each, both intrinsic source behavior and the application of stimulation from external or implanted devices. We will develop algorithms and tools for simulation (by which we mean making assumptions about anatomy and/or function and then using computation to predict measurements and outcomes) and estimation (by which we mean starting with measurements and a parameterized model of anatomy and/or function and using computation to estimate the values of the parameters). Our specific aims encompass three broad objectives, in concert with our own tech- nical expertise and research interests. The first objective is to provide algorithms and software for constructing and executing simulations and calculating relevant estimates from measurements. Our second objective is to engage biomedical scientists and engineers in using our software tools. Our third objective is to enable access to experimental and clinical data to support validation of results and comparisons across algorithms. Cutting across all three objectives is the growing need to quantify uncertainty as it affects all stages of simulation and es- timation, driven by the growing use of these techniques in both modern clinical medicine and modern biomedical research. As simulation and estimation transition into clinical medicine there is a growing need for comparison and validation of computational methods across experimental and clinical investigators. We have four specific aims. The first will develop and disseminate new algorithms and software for simulation and estimation for brain and cardiac bioelectric fields which stem from the intrinsic electrical activity of those organs. The second will also develop and disseminate new algorithms and software, but this time for bioelectric fields generated by manufactured devices, either implanted or at the body surface, such as cardiac defibrillators or devices used to modulate brain activity by injection of small currents. The third concerns designing software pipelines to help our collaborators build appropriately individualized computational models to use in the compu- tations of the first two aims, trading off between accuracy and time cost to create the models. The fourth aim will produce algorithms and software tools that calculate, describe, and lead to visualization of, the degree to which relevant aspects of our simulations and estimates are uncertain, in the sense that they are sensitive to unavoidable variability in our assumptions and random noise in our measurements. Taken together, our aims address essential needs for many of our collaborators, even those not concerned with bioelectric fields. Our success will reach beyond our immediate collaborators and impact a wide range of related biomedical research and science. Our past success and the close interactions we will pursue across the other parts of the Center, our Center's infrastructure, and the prominence and activity of our partners, all support our optimism that we can achieve our goals.