NIBIB Supports Modeling, Analysis, and Simulation in Clinical Medicine and Biomedical Research

By Margot Kern and Grace C.Y. Peng

When determining whether to open a blocked artery with a stent, cardiologists calculate the fractional flow reserve (FFR), a number that predicts the likelihood that a blockage is significantly impeding blood flow to the heart. Current methods for calculating FFR involve threading a catheter from the arteries in a patient’s groin or wrist all the way into the coronary arteries. It’s an invasive test, and one that costs several thousand dollars.

Recently, a team of surgeons and biomedical engineers has created an alternative method for determining FFR that doesn’t require catheterization. The method—called HeartFlow—uses anatomical data from a CT scan of a patient’s heart to create a realistic 3D multiscale computational model of the patient’s coronary arteries. Virtual blood, obeying the physical laws of fluid dynamics, runs through the arteries; the FFR can be calculated noninvasively at all points along the arteries. Three recent clinical trials indicate that HeartFlow shows promise for impacting current state-of-the-art diagnostic methods [1–3]. In the future, doctors could use the model to try out virtual stents, determining optimal size and location for stents to be placed in a patient.

HeartFlow is just one example of how computational modeling can help solve complex problems in clinical medicine and biomedical research. Biomedical modeling provides a powerful platform for improving the efficiency of scientific experimentation, data acquisition, knowledge transfer, and technology development. Furthermore, models can be used to test the safety and efficacy of medical devices and drugs in silico, reducing the dependency on human and animal studies and accelerating translation from basic science to clinical medicine. 

Support for Biomedical Modeling, Simulation, and Analysis Research

NIBIB—the National Institute of Biomedical Imaging and Bioengineering—is one of 27 institutes and centers that make up the US National Institutes of Health, the federal government’s premier biomedical research agency. The mission of NIBIB is to support the development of novel biomedical technologies to help diagnose, treat, and prevent illness.

Among many other focus areas, NIBIB provides support for quantitative research projects that contribute to its mission. These projects generally fall into three categories: 

1. Modeling: biomedical models and computational algorithms with potential clinical or biomedical applications. Research in this area includes mathematical, statistical, transport, network, population, mechanical, electrical, and electronic models applied to a broad range of biomedical fields. Particular emphasis is placed on multiscale modeling and the methods used to bridge scales. 
2. Analysis: mathematical, statistical, and signal processing methods for the analysis of complex biomedical systems, clinical diagnosis, and patient monitoring
3. Simulation: technologies for training and education in clinical practice and biomedical research, and simulation methods for emulating work flow, and understanding and predicting health and disease

Interagency Modeling and Analysis Group and the Multiscale Modeling Consortium

NIBIB coordinates the Interagency Modeling and Analysis Group, known as IMAG, which is made up of more than 80 program staff from ten government agencies who are involved in managing research programs in biomedical, biological, and behavioral systems; the programs have in common a need for novel modeling and analysis methods. From its inception in 2003, the group has provided an open forum for sharing updates on individual programs from the various IMAG agencies and for planning transagency activities that will have a broad impact on the communities served by IMAG.

Early in its existence, IMAG recognized that the modeling community was at the forefront of efforts to develop computational methods across the biological continuum. In addition, IMAG identified a strong desire among modelers to form multidisciplinary partnerships across varied research communities. This led to a 2004 solicitation: Interagency Opportunities in Multiscale Modeling in Biomedical, Biological, and Behavioral Systems. In 2006, the 24  awardees from the solicitation formed the Multiscale Modeling Consortium, which has grown to include more than 100 projects relevant to multiscale modeling. 

Funding announcements, meetings, workshops, tutorials, and other information relevant to the multiscale modeling community can be found on the IMAG Wiki page.  

How Can the SIAM Community Help Promote Biomedical Modeling, Analysis, and Simulation?

Although computational models are increasingly recognized as an important platform for understanding complex biological systems and facilitating technology development, some in the biomedical community continue to question their usefulness. A 2012 article in Biomedical Computation Review titled “Meet the Skeptics” explored the hesitancy of biologists and physicians to incorporate biomedical models into their work [4]. The article suggested that the issue stems in part from a combination of biologists who are intimidated by complex math and modelers who are dismissive of those not quantitatively trained. In addition, critics argue that modelers often don’t fully understand the biological conditions they are trying to model and, as a result, base their models on inaccurate assumptions. Some biomedical researchers, while acknowledging the promise of computational modeling, either believe that their biological problems are too complex for a model to be helpful or assume that they don’t have enough data to build an accurate model. Finally, a lack of well-known examples of biomedical models that have had a major impact on human health has led some to question their ultimate usefulness.

Members of the SIAM community can help assuage these concerns and become champions of computational modeling, analysis, and simulation in the following ways:

• Be prepared to describe and explain your models and analytical methods as simply as possible, working to make them accessible to potential users. 
• Explain that models are platforms for iterative research efforts, and not necessarily the end solution.
• Be prepared to explain how modeling can efficiently and economically drive scientific experimentation, data acquisition, technology development, and even policy. Use examples to show how modeling has the potential to accelerate translation from basic science to clinical medicine.
• Share success stories. Do you have or know of a successful biomedical model that has predicted outcomes that are difficult or impossible to obtain through experiments? If so, NIBIB wants to know about your story (which can be sent to [email protected]).
• Finally, members of the SIAM community are encouraged to apply for NIH funding and to participate in the NIH Peer Review process. Information on serving as a reviewer can be found here.