Professor
Bin Chen
Xi’an Jiaotong University, Xi’an, China

Title: The Integrated Imaging System Coupling Laser Speckle and Hyper-spectrum for Personalized Therapy of Dermatosis
Abstract: Vascular and pigmented dermatosis will influence patients’ health, both physiologically and psychologically. Laser therapy has become the golden standard, but total clearance rate is still quite low owing to lack of skin structure detection technique. In the present work, an integrated imaging system coupling laser speckle and hyper-spectrum was developed. Through this system, blood vessel diameter and depth can be detected and hemodynamics can be online monitored by laser specking imaging and inverse radiation method, while chromophore content including hemoglobin and melanin as well as thickness of epidermis can be measured by hyperspectral imaging, simultaneously. Laser-induced thrombus formation is the premise of desired thermal response of blood vessel, which had been related with tissue structure by our in-house code. In this way, multi-pulsed laser parameters can be optimized aiming to generate desired thermal response and achieve personalized therapy.

Professor
Bin Liu
Dalian University of Technology, China

Title: Personalized Computer-Assisted Medicine Based on Graphics and Image Processing
Abstract: The human pursuit of health is endless. Nowadays, countries all over the world have increased their investment in the field of health care. As an emerging discipline in the future, computer-assisted medical technology has attracted much attention. This is a deeply intersecting discipline. It not only includes computer technology and medical technology, but also requires the intervention of artificial intelligence technology, mechanical engineering, mechanics and other disciplines. In the past ten years, our research group has done some exploratory research on computer-assisted personalized orthopedic surgery (especially personalized orthopedic prosthesis modeling). In addition, we have also conducted some more challenging research in intelligent image processing and analysis. In future research, it is hoped that more disciplines can work together to truly solve clinical practical difficulties.

Visiting Professor & Co-Director
Feng Lin
Nanyang Technological University, Singapore

Title: High-Dimensional Complex Prior Knowledge Representation for Efficient Data Mining and Decision Making
Abstract: The core of artificial intelligence (AI) is knowledge representation and supervised / unsupervised machine learning (ML). Any application domain-specific AI systems must start with an abstract representation of its knowledge space of all dimensionalities, and a sampling dataset. We propose a generic knowledge representation scheme with its advanced AI algorithms for unsupervised clustering (a popular ML approach) of dataset. Based on comparative analyses, three major innovations will be: Uncovering the hidden distribution patterns of a high-dimensional and complex dataset, Evaluation of minimum bounding box of the dataset to maximize data mining resolution, and Unsupervised hierarchical clustering of complex data.

Distinguished Professor
Jyh-Cheng Chen
National Yang-Ming Chiao-Tung University

Title: Improvement of a quantitative analysis system for PET/MRI brain functional images in the treatment of Parkinson's disease with acupuncture
Abstract: Background and Purpose: Parkinson's disease (PD) is a neurodegenerative disease in old people whose pathogenic mechanism turns into the degeneration of substantia nigra of the midbrain, causing the degeneration of dopaminergic neurons in the putamen and caudate nucleus of the striatum, and the dopamine content in the synapses decreases and gradually loses the ability to act properly. Therefore, our purpose is to establish an automated quantitative analysis system for PET / MRI images, through the use of putamen and caudal ROI segmentation and MRI images to automatically perform quantitative analysis of PET assessment for PD therapy.
Methods: We used the open source SPM (statistical parametric mapping) software to co-register each PD patient’s MRI images and the T1 MRI template. Then use the SPM software module to spatially normalize the previously completed MRI images that have been co-aligned to set the clinical image’s spatial coordinate axis and voxel size (2 mm × 2 mm × 2 mm)…