MRI Infusion Pump Advisor Dr Naomi Chesler Biomedical

MRI Infusion Pump

Advisor Dr. Naomi Chesler § Biomedical Engineering § University of Wisconsin - Madison

Group Members § § § Ryan Augustine Gordy Lawrence Tim Eng Christine Weisshaar Megan Buroker Nate Gaeckle

Client § § § Dr. George C. Newman Department of Neurology University of Wisconsin Hospital

The goal of this design is to create a more time efficient pump that wastes a smaller amount of gadolinium and infuses with controllable precision. A prototype has been constructed that uses Bernoulli’s principle of ideal fluid flow. This prototype will be able to deliver a bolus injection, immediately followed by the infusion, while also having the ability to change the flow rates of gadolinium and saline throughout the infusion. This will save both time during the procedure, and money expended on gadolinium. The design is still in the beginning stages, and there a few modifications and additions to be made to have a device that can be used accurately in an MRI environment.

Background § Contrast agent gadolinium infused in blood stream to create diagnostic image. § Infusion pump holds syringe of gadolinium and syringe of saline rinse. § Pump connected to patient’s IV line. § Magnetic field requires materials be non-ferrous § Gadolinium is pre-packaged. – Waste $40. per bottle

Motivation § Our client desires a controllable device for bolus and infusion during MRI scans that is durable enough for daily use. Inaccurate flow rates and quantities can lead to diagnostic errors.

Client Requirements § Current MRI pumps are restricted in flow rate control and waste considerable contrast. Manual syringe loading costs time between bolus and infusion, leading to image inaccuracy. The client desires a pump that can control the flow rates of both contrast and saline, without having to constantly change the syringes. Due to the magnetic field, this pump must be made from nonferrous material.

Design Requirements § § § Flow rate ranges from 0. 25 to 5 m. L/sec Overcome venous pressure of 15 mm. Hg Decrease time between bolus and infusion Non-ferrous material. Bolus: deliver 10 -25 m. L +- 0. 5 m. L contrast at 3 m. L/s +- 0. 2 m. L/sec § Infusion: (within 0. 02 m. L/sec) – Deliver contrast at 0. 25 -0. 35 m. L/sec – Deliver saline at 0. 65 -0. 75 m. L/sec

§ Bernoulli’s Equation for steady Flow P 1 = P 2 + (1/2)ρQ 22(1/A 22 – 1/A 12) + ρgΔh § Bernoulli’s Equation for unsteady Flow P 1 = P 2 + (1/2)ρQ 22(1/A 22 – 1/A 12) + ρgΔh + ∫(dvs/dt)ds

Final Design § Follows Bernoulli’s Equation for Ideal Fluid Flow § Made for 250 m. L IV bags. § Pump made from polycarbonate and brass fittings. § Rubber Gasket used to seal the pump. § ~$5000 for final design.

Future Work § Addition of Ultrasonic flowmeters to IV lines. § Addition of pressure regulators between the gas canister and pump. § Integrate with computer controlling. § Obtain a more sensitive regulator to improve accuracy. § Construct stand to hold apparatus. § Add second pump for gadolinium. § Construct an adapter to change IV tubing.

References Mc. Master-Carr. (2003). Retrieved Nov. 3, 2003, from http: //www. mcmaster. com Medrad. (2003). Spectris Solaris MR injection system. Retrieved Dec. 4, 2003, from http: //www. medrad. com/systems-and-products/magneticresonance/spectris-solaris. html Newman, George C. Personal Interview. Sept. 12, 2003. Tummescheit, H. (2002). Modeling pitfalls: Some examples in modelica. Retrieved Oct. 17, 2003, from http: //www. control. lth. se/~hubertus/pitfalls. pdf