کنفرانس بین المللی فناوری و مدیریت انرژی

صفحه اصلی / نهمین کنفرانس بین المللی فناوری و مدیریت انرژی

Energy consumption optimization & Improving performance in CPU of the implantable cardioverter defibrillator with using design and hardware implementation of CNN neural network on Zynq

نویسندگان :

Alirea Keyanfar¹ Reza Ghaderi² Soheila Nazari³ Behzad Hajimoradi⁴ Leila Kamalzadeh⁵

1- دانشگاه شهید بهشتی 2- دانشگاه شهید بهشتی 3- دانشگاه شهید بهشتی 4- دانشگاه علوم پزشکی شهید بهشتی 5- دانشگاه علوم پزشکی ایران

کلمات کلیدی :

Energy consumption optimization in medical devices

چکیده :

Arrhythmias of the heart, including ventricular fibrillation (VF) and ventricular tachycardia (VT), must be diagnosed and treated promptly. In order to be effective, implantable cardioverter-defibrillators (ICDs) must promptly identify and treat ventricular tachycardia (VT) and ventricular fibrillation (VF). In order to optimize energy consumption and boost the performance of the boundary core of the implanted defibrillator device, this research aims to provide the practical ways of employing deep learning in the processing of heart electrophysiological data through the use of a single network. This network is CNN adapted for usage in cardiac pacemakers. The network has collected data from forty patients with cardiac arrhythmia and forty patients without cardiac arrhythmia who had ICD checkups over the course of eight months. This network is designed to be as easy and accurate as possible for recognizing VF and VT EGM signals. Because of the critical nature of energy use in implanted medical equipment, it is crucial that designers prioritize efficiency whenever possible. The optimal number of parameters can increase network speed in signal processing and arrhythmia detection and can also be useful in reducing battery consumption. Finally, the designed CNN network hardware was implemented. zynq chips have the ability to process in parallel and can be useful in increasing the processing speed, so zynq chips were selected for the hardware target. After the hardware implementation stage, it is possible to proceed from the IP Core produced to design other parts of the defibrillator in the Vivado software.