Implantable NMR Microprobe Design using a Machine learning approach
Résumé
In this work, we propose the use of an Artificial Neuronal Network (ANN) to achieve the design process of an implantable Nuclear Magnetic Resonance (NMR) microprobe. Before the micromanufacturing process, we need to optimize the microprobe i.e. micro-coil + transmission line and the connecting path (wire-bonding, underpass&vias or airbridge) to get an optimal performance (high Qfactor and Noise signal Ratio (NSR)). An ANN model is developed from the simulated dataset and used to predict the optimal microprobe performance for a working frequency and the used substrate.
Origine : Fichiers produits par l'(les) auteur(s)