User Based Design of Medical Devices for Translation from Prototype to Clinical Device

In this work, two devices currently used in a research capacity were updated based on user and clinical requirements. Consideration was given for final designs with high reliability, broad applicability, and minimal amount of required training or technical skill to operate. An implantable microdevice allows for accelerated in vivo testing of anticancer agents on human derived tumors. The current method of filling the reservoirs of the microdevice with anticancer agents consists of manual stuffing of the reservoirs until the hole drilled in the device appears full. This process is labor intensive to a degree that prohibits large scale application and produces variable volumes depending on the skill of the technician and the tolerances of the manufacturing process for the microdevice. We designed and tested a new method of creating standardized drug volumes that are easy to load into the microdevice. We machined a master mold of stainless steel that was used to cast a flexible mold of PDMS. The pouring in of heated anticancer agents produces small drug pellets of a reproducible volume. We validated the method with fluorescent release studies which tested the variability in the pellet volumes between wells compared it to the variability found between and within users in the previous method of manual filling. The new method led to a reduction in the coefficient of variation of the amount of drug in each well from 49.92% to 15.32%. Additionally, we compared the time to fill a device with the two different methods to get an evaluation of technician skill and time required in a scale up of the two methods. A linear peristaltic nanofluidic pump actuated by NITINOL wire is currently used to sample rodent brains with minimal scarring [49]. This pump is in prototype stage and therefore components regularly break and it is oversized at a volume of 584cm3, with the need to replace the batteries on a regular basis. Investigation of the breaking of the NITINOL wire pinpointed friction as the likely primary cause. A newly designed tubing holder improves the average lifetime of the NITINOL wire from 6.1E3 cycles to 6.6E4 cycles and the minimum breaking time at the tubing holder from 119 minutes to 455 minutes. The breaking of the tubing is investigated and recommendations for future designs include reducing the heat transferred to the tubing or using a tubing material with a higher tear strength, lower compression set, and/or higher yield strength. The battery size required is minimized by reducing the NITINOL power drain with lower resistance electrical connections, a modified voltage activation profile that prevents overheating of the wire, and an investigation of the design changes required to reduce the wire length. These modifications reduced the required battery capacity by a third and transform the pump into a wearable device size.S.M