• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

About FDA

  • Print
  • Share
  • E-mail

FY 2001 Computational Modeling


Continuing advances in computer technology now make computational modeling a powerful tool for evaluating a variety of problems for which the underlying mechanisms are understood yet analytic solutions are intractable. Product designers are making increasing use of such modeling for product development; OST is making increasing use of such modeling for product assessment. These techniques allow scientists to manipulate a wide range of variables without having to create a physical representation of each possibility.

A well-optimized complementation of clinical trials with computer modeling holds excellent promise for both reducing costs and increasing product quality. For example, computational techniques often provide the best available information about blood flow through cardiovascular devices such as blood pumps and heart valves. Small regions of stagnant flow become breeding grounds for thrombosis, a potentially fatal complication. The information derived from computational analyses can be used in design optimization to suppress these zones of stagnant flow beyond what can be detected via clinical trials.

The goal of OST's investigations is to develop techniques that will enable the analysis of products and provide basic insight into the roles played by individual variables on the final outcomes. There are two major areas of current investigation in OST's Division of Physical Sciences (DPS).

A Computer Model for Determining Velocities of Heart-Valve Leaflets

Key words: computer simulation, prosthetic heart-valve dynamics, heart-valve cavitation

The tip velocity of a closing heart valve leaflet is known to be a useful parameter for quantifying the cavitation potential of prosthetic valves. Unfortunately, measurements of the leaflet velocity are difficult to make, requiring sophisticated techniques and equipment. Full numerical simulations of the closing process are likewise very complicated. A simple procedure for determining leaflet velocity in terms of easily measured quantities is highly desirable.

In the case of rigid cavitation testers, approximations can be made which make a simple mathematical model feasible. In rigid cavitation testers and pulse duplicators, impulsive motion at the boundary (often an accelerating piston) produces large pressures and temporal gradients in fluid velocity near the valve, but relatively small velocities. Under these "impulsive motion" conditions, the pressure field satisfies Laplace’s equation, i.e., determining the pressure field is essentially reduced to solving a low-frequency scattering problem. Leaflet motion is calculated by first solving Laplace’s equation in the geometry of the test apparatus, using stationary leaflets at various angles. The resulting pressure "scattering coefficients" are then stored and used in a simple rigid-body analysis to determine the leaflet motion. Additional input to this rigid-body analysis is the ventricular pressure, which is easily measured, along with the leaflet properties.

The impulsive-motion model just described was used to analyze data from 12 tests of Edwards-Duromedics valves at the cavitation threshold. Experiments were conducted in the FDA cavitation tester, as part of a 1994 round-robin study. The average calculated closing time of 0.021 (+/- 0.003) seconds compared well with the mean time of 0.019 (+/- 0.005) seconds determined from videotapes. The mean computed tip speed was 2.6 (+/- 0.3) m/s, in agreement with published values. Application of the model to other valve types is under investigation.

Computational Studies of Vascular Grafts

Key words: vascular grafts, cardiovascular devices, artificial organs, computational studies

After nearly 40 years of development, devices to replace diseased arteries still do not perform perfectly, especially in small diameters. Failure of small-diameter vascular grafts usually occurs by a combination of thrombosis and intimal hyperplasia (overgrowth of the attaching artery) which occludes the vessel. Intimal hyperplasia has often been noted to occur preferentially at the downstream junction between the vascular graft and native artery. This suggests that a flow-related mechanism is at least partly responsible for the failure. Laboratory experiments in transparent models have shown that small differences in the stiffness or diameter between the vascular graft and native artery can cause the trapping of tiny particles used to simulate blood cells at the downstream junction. If this happens in patients, the trapping of blood cells can cause agglomeration of cells and release of biochemicals which might cause the observed thrombosis and intimal hyperplasia.

This complex situation is now being modeled in a computer simulation, where greater control over the experimental parameters and more detailed examination of the results are both possible. Preliminary results show that a diameter/stiffness mismatch at the downstream junction does indeed enhance the concentration of a dissolved constituent. If confirmed, this research will provide additional support to the hypothesis that minor geometrical abnormalities can affect the proper healing of a vascular graft through a flow-related mechanism. This project also demonstrates the usefulness of computer simulations in developing cardiovascular devices. Such simulations are expected to gain importance in the device applications that the Center receives in the future.

Predicting the Intraocular Pressure Following Pneumatic Retinopexy

Key words: detached retinal, perfluoropropane, sulfur hexaflouride

CDRH received a PMA for a new use of perfluoropropane gas retinal tamponade. In this technique a gas is injected to push the detached retina back into place. Once the retina is in position, cryotherapy is typically used to weld the retina in place for subsequent regeneration of the pigment epithelium. In the eye, blood gases diffuse into the bubble causing expansion, followed by desorption of the gases over a period of several weeks. During this period, elevations in intraocular pressure are possible, which could have long term adverse affects on the patients’ vision. OST developed a mathematical model of the gas diffusion of the injected gas, blood gases, and pressure dynamics to predict conditions that could result in elevations of intraocular pressure. The model predicts the expansion and persistence of the gas, as well as injected gas compositions that minimize expansion and pressure elevation effects. OST was also able to correlate molecular volume of different candidate materials with the diffusion rate in the eye.

The model has been submitted for peer-reviewed publication in Current Eye Research and a manufacturer has expressed interest in marketing the model to outside parties. This model will allow ODE to substantially reduce animal testing. The model can accurately track the non-expansive behavior of tamponade gas - air mixtures. Once the model is correlated in the rabbit, it can reduce the number of experiments needed to establish the persistence in the eye.

Predicting the Shelf Life of Dialyzers

Key words: polymer degradation, shelf life, dialyzers

OST developed a Monte Carlo model to predict the chemical degradation of dialysis membrane materials in storage. The model tracks changes in a population of molecules representative of the dialyzer membrane. The degradation process calculation is a two-step process: first the model uses a random number to select an individual polymer molecule out of the population; then a second random number is used to select the site on the molecule for the degradation reaction. After the reaction, the resulting fragment molecules are redistributed into the population, and population average properties are recalculated. The model has been applied to cellulose acetate and polyethersulfone materials. A shelf life of 2 years was identified for cellulose acetate dialyzers based on animal test results. Polyethersulfone work is still in the preliminary stages, since this polymer has a more complex degradation mechanism.

New methodology to predict polymer degradation will help CDRH and device manufacturers’ predict chemical degradation effects based on chromatographic data acquired in short term experiments.

Performance Characterization of Accelerator Target Systems for Neutron Capture Therapy

Key words: radiation therapy, neutrons, cancer

Boron Neutron Capture Therapy (BNCT) is an investigational therapy for brain tumors that are refractory to current therapies. Boron Neutron Capture Synovectomy (BNCS) has been proposed for treatment of rheumatoid arthritis. The continuing development of proton accelerator technology will result in greater clinical application of these therapies. Although BNCT is an active area of clinical investigation, dosimetric methods have not been standardized, making it very difficult to compare results of clinical trials to conventional treatments to assess efficacy. This project will help to standardize the methods used for BNCT dosimetry.

OST is studying neutron source design for clinical applications using computational modeling of neutron transport. During FY 2001, the models were extended to include the detailed physics of neutron sources available to OST. A model of a human phantom was used for analyzing radiation doses to several organs. OST developed detailed models of the neutron moderators used for its experiments. The researchers are developing in vitro models of BNCT using cultured brain tumor cells loaded with p-Boronophenylalanine (BPA). OST confirmed that the brain tumor cells concentrated the BPA to three times the external concentration and has signed an interagency agreement with DOD to use DOD facilities for experiments. During FY 2002, a significant computational effort will be devoted to the design of the shielding cave for the planned OST facility and scientists will continue dosimetric measurements and biological experiments on the DOD neutron source.