Keynote Speakers

Prof. Kuo-Sheng Cheng
National Cheng Kung University, Taiwan

Prof. Kuo-Sheng Cheng received his BSc. and M.Sc. degrees both in electrical engineering in 1980 and 1982 from the National Cheng Kung University, Tainan, Taiwan, and the M.S. degree in biomedical engineering in 1988 from Rensselaer Polytechnic Institute, Troy, NY, USA. In 1990, he received a Ph.D. degree in electrical engineering from the National Cheng Kung University. Currently, he is a Professor at the Department of Biomedical Engineering and an Adjunct Professor at the Institute of Oral Medicine at National Cheng Kung University. He also serves as the Director of the Department of Medical Engineering at National Cheng Kung University Hospital and the Director of the Engineering and Technology Promotion Center financially supported by the National Science and Technology Council, TAIWAN. He is also the Chair of the IEEE EMBS Tainan Chapter. His research interests include medical image processing and analysis, electrical impedance tomography system development and applications, and biomedical instrumentation and measurement.

Speech Title: "The Applications of Deep Learning in Separation of Heart and Lung Impedance Images in Electrical Impedance Tomography"

Abstract: Electrical impedance tomography (EIT) is a noninvasive medical imaging technique that measures the impedance of tissues by applying low-amplitude electrical currents and measuring the resulting voltages. The impedance distribution inside the body can then be reconstructed by solving an inverse problem, which provides useful information for diagnosing and monitoring a variety of conditions, such as respiratory and circulatory functions. One of the challenges of EIT in chest applications is separating the impedance signals from the heart and lungs. This is difficult because the two organs are close together and their impedance signals are mixed. However, deep learning has emerged as a promising approach to solving this problem. Deep learning models can be trained on EIT data to learn the features that distinguish between heart and lung impedance. Once trained, these models can be used to separate the two signals and produce more accurate EIT images. In our recent study, we proposed a novel semi-Siamese U-Net architecture for heart and lung impedance image separation. This architecture is based on the state-of-the-art U-Net, with modifications and extensions to improve performance on EIT image separation.