Home

Special Session 4: GRAPHIC: Graph-Based Reasoning and Analysis of Physiological and Healthcare Information for Explainable Clinical Decision Support

• Brief Description:
  Artificial intelligence is increasingly used to analyse medical images, bioacoustics recordings, physiological waveforms, and multimodal clinical data. However, many systems still present their outputs as isolated labels, probabilities, or static heat maps. These outputs can be difficult for clinicians to interrogate and relate to the underlying clinical evidence. Graph based representations offer a promising way to connect anatomical locations, signal segments, physiological events, imaging findings, and temporal relationships within a single interpretable framework.
  This special session will bring together research on graph learning, visual analytics, medical image and signal processing, multimodal fusion, and human centred clinical AI. It will examine how model outputs can be linked directly to source evidence, including images, waveforms, spectrograms, anatomical maps, and replayable audio. Particular attention will be given to systems that enable clinicians to inspect abnormal events, compare signals across locations or time, understand uncertainty, and verify the basis of an AI generated recommendation.
  Applications may include respiratory disease assessment using digital auscultation, cardiac sound and ECG analysis, sleep disordered breathing, medical imaging, remote monitoring, and other clinically relevant multimodal data. The session welcomes theoretical advances, new algorithms, interactive systems, datasets, and clinical evaluation studies. Its focus aligns with ICIGP themes in medical image analysis, deep learning, multimedia processing, visualisation, user experience, and interface design.
  
  
  Session Organizers:
  Assoc. Prof. Wei Quan, University of Lancashire, UK
  Prof. Zhen Lin, University of Sanya, China
  
  
  The topics of interest include, but are not limited to:
  • Graph neural networks and representation learning for clinical data
  • Clinical knowledge graphs and evidence graph for decision support
  • Spatial and temporal graph modelling of physiological events
  • Explainable and trustworthy AI for medical image and signal analysis
  • Visual analytics and interactive interfaces for clinical AI
  • Medical image analysis and multimodal image signal fusion
  • Visualisation and replay of bioacoustics signals and spectrograms
  • Human in the loop learning and clinician AI interaction
  • Uncertainty, calibration, signal quality, and transparent model reporting
  • Privacy preserving and federated graph learning for healthcare
  • Datasets, benchmarks, reproducibility, and real-world clinical evaluation
  • Applications in respiratory, cardiac, sleep, and remote health monitoring
  
  
  Submission Method
  Submit your Full Paper (no less than 5 pages) or your paper abstract-without publication (200-400 words) via Online Submission System, then choose Special Session 4 (GRAPHIC: Graph-Based Reasoning and Analysis of Physiological and Healthcare Information for Explainable Clinical Decision Support)
  Template Download: (MS Word Template) (LaTeX Template)
  
  
  Introduction of Session Organizers
  
  

  
   

  Assoc. Prof. Wei Quan
  University of Lancashire, UK

   

   

   

  Dr Wei Quan is an Associate Professor in Computer Vision and Machine Learning at the University of Lancashire. His research covers computer vision, biomedical signal processing, machine learning, visual analytics, digital health technologies, and non-contact physiological sensing. He has published more than 50 peer reviewed papers and has received several best paper awards, including awards at ICPRAM 2015, ICIGP 2021, and IFSP 2025 & 2026.
  His current work investigates explainable visual decision support for multimodal clinical signals, with applications in digital respiratory auscultation, cardiac sound analysis, and sleep disordered breathing. He collaborates with UK NHS clinicians and industry partners on clinical data collection, algorithm development, training tools, and the translation of signal processing and AI methods into healthcare practice.
  
  
  
  
  

  
   

  Prof. Zhen Lin
  University of Sanya, China

   

   

   

  Professor Zhen Lin is a Professor and Dean of the School of New Energy and Intelligent Connected Vehicles at the University of Sanya, China. Her research interests span autonomous driving, new energy vehicle technologies, intelligent connected vehicles, unmanned systems, agricultural robotics, artificial intelligence, and smart agricultural equipment.
  Her current research focuses on autonomous systems and intelligent equipment, including autonomous navigation and path planning, intelligent control, sensor fusion, embedded systems, and unmanned agricultural machinery. A key aspect of this work is the intelligent interpretation and integration of heterogeneous sensing information for perception, navigation, monitoring, and decision-making. Her broader research interests encompass computer vision, image and graphical information processing, multimodal sensing, AI-based perception, and interpretable representation of complex sensor information. Professor Lin has led and contributed to more than 10 national, provincial, and local research and education projects, including projects funded by the Hainan Provincial Natural Science Foundation and the Sanya Science and Technology Programme. She has published over 20 academic papers, including SCI-, EI- and core-journal publications, authored an academic monograph, and holds nine granted patents and two software copyrights. Her research has practical applications in smart agriculture, autonomous equipment, and new energy vehicle technologies.