Prof. Dr. Professor of College of Medicine and Biological Information Engineering, Hunnan Campus of Northeastern University, China Director of the biomedical electronics institute in Sino-Dutch Biomedical and Information Engineering School of Northeastern University; Professor at the Key Laboratory of Medical Image Computing of Ministry of Education, China; Chairman of theory and education professional committee of China Medical Informatics Association; Vice chairman of TCM diagnosis information committee of Chinese medicine information society; Senior member of IEEE Society; Senior member of Chinese Society of Biomedical Engineering; Member of the editor board for many international journals such as Physiological Measurement, Biomedical Engineering Online, Computers in Biology and Medicine and so on.
Title:Noninvasive monitoring of aortic pressure wave based on peripheral pulse waves
Abstract: Blood pressure wave can reflect the status of cardiovascular system. In comparison to peripheral arterial pressure waves, aortic pressure wave is more efficient in predicting cardiovascular events. However, the direct measurement of aortic pressure wave is invasive, complex, expensive, and has some high risks. Therefore, the non-invasive estimation of aortic pressure wave has attracted many concerns. The transfer function methods have been studied for a long time and have achieved many successes. Currently, the transfer function has changed from population-based method to individualized method. There are many ways to individualize the transfer function of cardiovascular system. This presentation will show some work of our group to individualize the transfer function by adaptively adjust the transfer function according to the systolic blood pressure acquired from the arm cuff, considering the time-variant characteristics of cardiovascular system using blind identification technique, combining with 0D (Windkessel) models, combining with temporal convolutional network. Furthermore, the advantages and disadvantages of these methods will also be discussed.
Associate Professor Polytechnic Institute of Portalegre, Technology and Management School, Portugal Head of the Electronics and Instrumentation Laboratory, Syllabus/Course Coordination, Teaching and Research, Vice-President of the Pedagogical Council
Title: Numerical Quantization of Neural Networks
Artificial Intelligence (AI) and its usage are becoming quite regular and spread daily. While applications such as chatbots, recommender systems, or spam filters rely on massive network structures that are deployed on remote servers, when it comes to natural language processing or healthcare monitoring, it is pretty crucial that AI runs its algorithms on a smartphone or a wearable device, both with memory and power restrictions. Also, running AI algorithms on edge or on the device can solve many issues due to privacy concerns or bandwidth limitations. This keynote will discuss the challenges in transforming numeric quantities to low-bit-width representation and the numerical error that arises.
© 2024 All Rights Reserved. CMME2024