From July 14 to 17, 2025, a six-member graduate team from the Institute of Advanced Medical Engineering and Equipment, School of Instrument Science and Engineering, Southeast University, will attend the 47th IEEE International Conference on Engineering in Medicine and Biology (EMBC 2025) at the Bella Center, Copenhagen. Founded in 1957, EMBC is one of the most influential conferences in biomedical engineering. The 2025 edition, co-hosted by the Technical University of Denmark and the University of Copenhagen, focuses on “Engineering Medicine, Innovating Healthcare” and has attracted more than 2,300 participants worldwide. Ahead of the trip, the SEU team held an online rehearsal within the school to refine their presentations and exchange feedback with faculty and peers.
Rong Xiao delivered a presentation entitled "ECG Signal Generation Using Variable β-Conditional Variational Autoencoder" and engaged in discussions with Dr Chang Yan，revising details presented in the slides.
Zhaoyang Cong delivered a presentation titled “Hierarchical Multi-Scale Feature Fusion Network for Multi-Center Major Depressive Disorder Classification with T1-weighted MRI.” The report proposed a novel 3D hierarchical multi-scale feature fusion neural network, which integrates multi-level gray matter and white matter features to achieve efficient multi-center classification of major depressive disorder based on T1-weighted MRI.
Lukai Pang delivered a presentation entitled “A Multimodal Data-Driven Assessment System for Autism Spectrum Disorder in Children” at the 47th IEEE EMBC in Copenhagen. She introduced a mobile platform integrating EEG, ECG, speech, facial expressions, and clinical scales for objective ASD assessment in children, and discussed future extension to younger age groups. She also exchanged views with experts on its potential as a low-cost, data-driven early screening paradigm.
Xiangyu Zhang delivered a presentation titled "Multi-Scale Convolutional Attention Enhanced Residual Network for Cuff-less Blood Pressure Prediction", which introduced a novel residual network framework integrating dynamic convolution and multi-scale attention mechanisms for accurate blood pressure estimation using PPG and ECG signals. The presentation also included a discussion and exchange with faculty members.
Yutong Zhang presented "A Disturbance Suppression Framework Based on Online Prediction for the Dynamic Measurement of Skin Temperature," introducing a framework that refines parametric methods in heat conduction analysis. Subsequently, she engaged in discussions with Professors Yang and Xing.
Zifei He introduced a poster, “Sensor-skin Interface Design for the Improvement of Reflective Photoplethysmography Signal Quality on Flat Skin Surfaces”. This work presented low-cost design considerations that improves the raw signal quality of reflective PPG. Major considerations were raised sensing surface and all-in-one electrode. The results show that the proposed device outperforms the conventional device at PPG signal.
