Laurine Kolly
Drug delivery systems (DDS) were engineered to overcome the drawbacks of more common methods such as oral pills and intravenous injections. Implantable capsules that release anesthetic drugs locally for pain management after Total Knee Arthroplasty (TKA) are currently in development within the LMIS1. The capsule will be implanted at the site of the wound during the operation and will be controlled wirelessly through resonant inductive coupling, enabling a local release of analgesics. Within this framework, this project aimed to investigate the interaction of electro-magnetic elds with the tissue that separates the transmitting coil of the controller from the implanted capsule. The main aspects that were studied consisted of the penetration of the magnetic eld and the Specic Absorption Rate (SAR). As the receiving coils on the reservoirs of capsule will each resonate at specic frequency between 300 MHz and 1.1 GHz, the aforementioned aspects were assessed at the two ends of the frequency spectrum. This was performed through the means of Finite Element Analysis, using COMSOL simulation software. To do so, models with three dierent types of media were built: air as a baseline, multi-layered tissue as an anatomically representative model, that reproduced remote power transfer to a capsule implanted at the level of the knee, and water as an approximation of the multi-layered tissue model. As a result, it appeared that, at the higher frequencies, the magnetic eld is subject to considerable attenuation, such that the power input necessary to generate sucient eld intensity at the level of the capsule in a multi-layered tissue model must be increased from a baseline of 5 W, to 10 W at 300 MHz and to 50 W at 1.1 GHz. The analysis of the SAR revealed that average peak SAR values reached are higher than the safety limit (in an uncontrolled environment) at a frequency of 1.1 GHz, and that the SAR is proportional to the power input. This implies that there is a trade-o to increasing the power input to adjust the intensity of the magnetic eld that could potentially be resolved by modifying the circuit design. Furthermore, it was concluded that approximating tissue with water may cause for the magnetic eld intensity as well as the SAR to be underestimated, due to the diering dielectric properties and to the density of water being lower than that of skin or tendon. Thus, a more anatomically accurate multi-layered tissue model should be favoured.
2020
