Modeling a Prescription for the Future: Faster, Safer Delivery of Therapeutic Substances.
Dr John Kalafut MEDRAD's Innovations Group
Computer modeling has proven its value throughout the design, development, and deployment of new products. Such is the experience of John Kalafut, a principal research scientist in MEDRAD´s Innovations Group (Indianola, Pennsylvania) who comments, “COMSOL Multiphysics has accompanied me throughout my career at MEDRAD, starting with systems engineering and today in R&D, even for products that won´t be available for many years”. Kalafut´s experiences as a medical engineer show how multiphysics modeling is useful for solving an exceptionally wide range of problem.
MEDRAD, with sales of roughly $500 million, manufactures, sells, and services medical devices for diagnostic imaging and therapy in three main areas: cardiovascular diagnosis, magnetic resonance imaging (MRI), and computed tomography (CT). The company has 1700 employees worldwide, and physicians around the world use the company´s products for more than 20 million medical procedures each year. One of the company´s core competencies is intravascular fluid delivery such as supplying exact doses of medicine or contrast agents. The Innovations Group investigates novel technologies, business opportunities, and clinical applications to continue the company´s >15% growth rate.
Five engineers in the company use COMSOL software for a variety of tasks. Comments Kalafut, “COMSOL Multiphysics is a natural choice to support us during the investigation of concept feasibility, IP due diligence, and in research. This is a very powerful tool for the corporate biomedical engineer in research and development. Its true multiphysics capabilities mean that ‘the sky’s the limit´ in terms of what we can tackle. COMSOL Multiphysics allows for the quick investigation of complex interactions-and a very affordable price”.
Figure 1a: The Stellant CT imaging system provides the ability to inject contrast material and saline at a measured ratio through a simultaneous plunger motion. Modeling helps determine the optimal rate of delivery.
From a modeling perspective, much of the firm´s research deals with the most efficient yet safest way to deliver diagnostic fluids into a patient´s body. And while fluid dynamics plays a crucial role in such studies, these models sometimes also involve heat transfer, electrostatics, chemical engineering, electromagnetics, and other physics.
Finding the best peak-enhancement curve
Speed of delivery plays an increasingly important role in improved CT scanners that allow for the acquisition of volumetric scans of the entire body in just seconds (Figure 1a). To achieve the superb diagnostic images possible with modern CT scanners, the injection and delivery of the contrast material must be synchronized with the imaging procedure. One benefit of new CT scanners is that because the imaging takes a shorter amount of time, the total dose of contrast material can be reduced. The timing window is short, but in that time doctors want to make certain they have good insight into how the material travels throughout the body.
Figure 1b: A simpliied model of the circulatory system from the point of entry for the contrast material leading to the heart. With this exploratory model, the engineers can determine which conditions give the optimal distribution.
Fast injections alter the enhancement profile and, because the flow is transiting so quickly through the vasculature, the delivery peak is sometimes not well synchronized and the patient must be reinjected and scanned again. In addition, each patient presents a different flow profile (time-density curves), which complicates rational material-delivery schemes. The key question becomes how much contrast material must you inject to get good images of the blood vessels and the heart? At what rate? How long should the injection last?
Figure 2: 2D model of a contrast material mixing with blood that allows for investigations of the dispersion of the material.
As is the case with many systems developed at MEDRAD, here a major goal is to get the maximum amount of contrast material into the bloodstream and heart as quickly as possible. Researchers want to study the dynamic forces that arise from the insertion of a viscous fluid through a tube at rates from 0 to 6 or 7 ml/sec.
To address these questions, the company is developing smart injection systems where a doctor first performs an identification injection, and the system then determines the proper amount of contrast material. In the research and feasibility phase of this technology, a model of the human body would be preferable to benchtop in-vitro investigation or animal models. However, it would be impractical to do a full finite-element model of the entire body, so Kalafut and his team concentrated on the vessels going from the point of injection to the heart, so as to determine what happens to the drug on its way there (Figure 1b).
They used COMSOL Multiphysics and the Chemical Engineering Module to better understand the early-time dynamics of the injection event. They coupled the Navier-Stokes application mode with the Convection & Diffusion application mode to gain insight into the distribution of the contrast agent through the peripheral vasculature into the heart. The early phase of the contrast's distribution in-vivo is difficult to determine non-invasively and is crucial for understanding the dynamics of contrast injection and propagation.
Knowledge gained by numerical modeling and simulation lends credence to assumptions made in global, compartmental models of the contrast material distribution after injection. One example of this model´s application is in understanding the transit time from the injection site, typically a large bore angiocath needle inserted into an arm vein, to the heart. In some patients, due to possibly diseased venous systems, blocked veins, or even the position of the arm during injection, not all of the injected contrast material arrives as a well-defined bolus. Rather, the contrast material can become dispersed and arrive over a period of different times. Such a situation makes predictive control and contrast material delivery difficult, if not impossible to conduct.
The COMSOL models are playing a role in discovering which factors influence the dynamics of the contrast material. Because it is injected at rates much higher than typically encountered in healthcare (2-8 ml/s), an understanding of the processes at the injection site is also an area ripe for investigation with COMSOL Multiphysics. Figure 2 displays a 2D simulation depicting the injection of highly viscous fluid into a large bore "blood vessel." The results from this simulation are coupled with a material and physical model to better understand the relationship between injection rate and the dynamics at the needle-puncture site. This knowledge could ultimately aid clinicians when assessing the likelihood of an injection site failing (an uncommon but aggravating situation) during the administration of contrast material. This model leads to better use of the material, better scans, and better diagnosis.
Catheter shouldn�t damage blood vessel walls
Figure 3a: The Vanguard DX catheter allows for a very uniform distribution of contrast materials. Laser-drilled holes or slits force the contrast material to be transported radially from the catheter.
In addition to designing vascular-injection systems, patient-monitoring systems, and MRI surface coils, MEDRAD also designs and sells diagnostic angiographic catheters. A particularly novel device is the Vanguard DX catheter (Figure 3a). Its nozzle design in the diffusion tip allows for a more uniform distribution of injected contrast material (the fluid used for creating a contrast during imaging) compared to a traditional end-hole catheter, which also tends to cause the contrast material to stream from the exit hole with high velocities. The Vanguard DX catheter reduces the reaction forces associated with contrast material streaming from the nozzle and therefore minimizes the likelihood of the catheter jerking against the blood vessel walls and damaging them.
Here a crucial question arises: What is the ideal configuration of holes or slits around the catheter tip to optimize fluid delivery while preventing a structural deflection? The researchers used COMSOL Multiphysics (coupling forces from laminar flow with a stress-strain analysis) to model the fluid-structure interaction occurring in the catheters with various hole configurations, geometries, and flow patterns (Figure 3b). Kalafut relates that, "one of our intern students, Ai Pi, an undergraduate bioengineer at Case Western Reserve University, generated many configurations of hole designs in different fluid regimes. We used these results to limit the number of benchtop models the mechanical engineers needed to fabricate and to help determine the feasibility of new ideas without needing to develop too many prototypes."
Figure 3b: An indication of the delection of a catheter tip similar to that in the Vanguard DX catheter with a series of holes (left) and slits (right) around its circumference in the middle. In this case, slits result in less delection and therefore less chance of damaging arteries.
Shielding a communications link
One of the company´s first serious experiences with finite-element modeling-and where they found that comsol Multiphysics makes modeling easy-came when fixing a problem that arose in an upgraded MRI contrast-injection system. That system uses an infrared link to communicate between the injection stand in the treatment room and the operator panel in a glassed-in control room (Figure 4a). Component obsolescence-an integrated circuit phototransducer taken out of production�required redesign of the communications link. The replacement part, however, brought with it some new problems; it was failing because it was more susceptible to electromagnetic interference from the MRI scanner and voltages inherent in the system.
Figure 4a: An MRI contrast-injection system consisting of a remote control panel that interfaces to the injection head located in the treatment room through a wireless link. A new phototransducer experienced electromagnetic interference and required shielding that was optimized using COMSOL Multiphysics.
Figure 4b: The COMSOL Multiphysics model of a phototransducer along with its shielding. The results show the E field around the inal geometry. The goal was to shield the openings from an external E field greater than 20 V/m.
The task then became one of determining how to reduce the effects of the electromagnetic fields on the phototransducer. Using COMSOL Multiphysics, Kalafut quickly replicated the problems that service engineers saw and concluded that a shielding structure around the transceiver structure would be a quick, cheap solution (Figure 4b). He explains, "we had to determine how thick the shielding should be and what shape it should have to do an adequate job, but also be as small as possible and fit in the existing equipment." Using COMSOL Multiphysics he set up a parametric solver to examine various geometries. "This approach saved us weeks of benchtop testing and technician time in building prototypes. In addition, the 3D plots we generated were quite useful in communicating the value of modeling with other engineers and with our management, who thereby first came to appreciate the value of virtual prototyping."
Optimizing the delivery of contrast material
Another interesting task MEDRAD´s R&D group has is to investigate technologies that might not be commercially viable for five to ten years so as to ensure the company´s future growth. One such area will be sending not only medicines or diagnostic fluids into a patient´s bloodstream but also living cells. Here the study of rheology and the geometry of the fluid path is critical because the delivery system cannot generate high shear forces, or else the delivery process can destroy the biologic material being injected.
Figure 5: Model of a syringe and needle with the velocity field distribution. The designers look for shear forces and use the model to help determine how to control them.
Marty Uram, a senior research scientist at MEDRAD Innovations, has been using COMSOL Multiphysics to model the fluid path between a vial containing the biologic agents (such as autologous stem cells) and the tissue into which the agent is injected. COMSOL allows for quick analysis of multiple geometries when attempting to trade off between laminar, high-velocity flow fields and regions of flow shear within the disposable tubing set and needles.
One of the very first questions to address is what kind of syringe will be necessary. Here it is important to model the transition from the syringe to the catheter. The idea is to create a flow transition that has no sudden changes in the flow path. The model in Figure 5 examines what changes arise in the drug fluid path due to changes in the outlet of a syringe design. Areas being investigated include: optimum configuration of holes in needles inserted into tissue, critical regions of shear using 3D Non-Newtonian fluid simulations, and insight into internal velocity fields not easily obtained with other methods. MEDRAD Innovations has recently deployed a Particle Imaging Velocimetry (PIV) system for measuring and quantizing fluid fields in the lab. The PIV system is being used to confirm the COMSOL simulation results.
Read the research paper at:
www.comsol.com/academic/papers/1020