Snaking pacemaker leads into a misfiring heart is not as easy as skilled surgeons make it look. So to model the problem and examine solutions, engineers with Boston Scientific found that the easiest way to model a flexible body was with rigid-body software, in this case Adams motion simulation software from MSC.Software, Santa, Ana, Calif. (mscsoftware.com). Rather than thousands of finite elements, a one meter pacemaker lead can be modeled with about 100, 1 cm segments connected by beam elements.

The simulation task is complicated by many small forces that come from the lead contacting veins, stylet (outer cover) friction on leads, and contact between other internal components. “The nonlinear FEA software we first tried took too much time to generate models and was inefficient for motion dynamics studies,” says principal mechanical engineer Jeff Bodner at Boston Scientific's CRM Div. Other alternatives for studying lead performance include animal models, which are expensive and come with ethical issues. Plastic or rubber models allow measuring forces but not motion, and how close either comes to reality is uncertain. What's more, testing different variables, such as the effect of a change in tubing stiffness on a physical heart model is time consuming and expensive.

“Computer simulations let us change parameters and quickly evaluate different configurations,” says mechanical engineer Adam Rivard. For example, testing an extendable and retractable pacemaker-lead implant begins with a surgeon bending the stylet into a curve to aid in its positioning. The outer cover stiffens the floppy lead so surgeons can push it against the target heart tissue and rotate a coil within the lead. The coil turns a mechanism that rotates and extends a helix, actually a miniature corkscrew which is the electrode, into the tissue.

“The software lets us bend the stylet into a curve just like a surgeon would,” says Bodner. “Also, we can simulate the cutting action of the helix and track its penetration into soft tissue. The torque needed to drive the helix into tissue is derived from an experiment and is used in the computer model. The model is so accurate that if we partially back the helix out and start again, the torque stays low until the helix reaches the point where the cutting previously ended.”

The software company helped build macros to reduce time spent modeling the leads. Without the macros it would take days to construct a model. The software now generates a working lead model in less than 30 minutes.

“Most of that time is spent thinking about which variables to consider and their corresponding values. The actual time to execute is extremely short. Now we spend more time thinking about the problem, not building the model. “

One complication from using motion-simulation software was contact between coaxial layers of the lead. The innermost stylet wire, inner coil, and outer assembly touch and interact with each other. Another custom routine written by the software developer simulates user-defined contact between lead layers. The routine replaces existing Adams logic, letting the model interact as physical leads do.

“We expect that combining Adams with Marc, MSC's nonlinear FEA software, will provide a multidiscipline code capable of opening further opportunities to model flexible leads and soft tissue,” says Bodner.