|< BK21 플러스 BEST 정보기술 사업단 세미나 개최 안내 >
개최일시 : 2017년 03월 14일 화요일 16:00 ~ 18:00
세미나 제목 : (1) Ultra-High B Diffusion Imaging of Cervical Spinal Cord
발표초록 : A nerve bundle consists of three major compartments for water MRI; intra-axonal (IA), myelin, and extra-axonal (EA) spaces. It is a straightforward to null the signal from the myelin space in diffusion-weighted image (DWI), particularly in clinical MRI system, by using a long echo-time (> 50 ms) for the short T2 relaxation time (~ 10 ms) of the myelin water. Therefore, IA and EA waters contribute the signal in the conventional DWI. An axon is surrounded by myelin (tens of lipid-bilayers) to protect and speed up the signal transmission. This electrical insulator, myelin, is also a barrier for the water transport between IA and EA spaces because of its hydrophobic side, and restricts the IA water movement within an intra-axonal diameter ( ~ μm) in the direction radial (perpendicular) to the fiber, while EA water molecule can persist its movement across many axons. In pathologic condition, the neuro-signaling is interrupted as a combination of axonal damage and demyelination. Although the damage in the axonal space may not be reversible, myelin damage can be repaired. Demyelination induces the increases in the water exchange between IA and EA space, and the IA water diffusion is no longer restricted within an axonal diameter.
These properties are used to evaluate the pathologic change in the spinal cord injury, using a ultra-high-B (UHB) DWI. To validate our assumption, we have developed, (a) DW-stimulated-echo to allow the water exchange with a long period without significant loss of signal, (b) numerical Monte-Carlo simulation to understand the meaning of the UHB-DWI signal, and (c) dedicated phase-array RF coils for cervical spinal cord. Our preliminary data will be presented using these tools for evaluating the pathology in the cervical spinal cord of a few patients with MS. This imaging method is potential for other spinal cord diseases, including cervical spondylotic myelopathy (CSM), multiple-sclerosis (MS), and ALS (Amyotrophic Lateral Sclerosis).
세미나 제목 : (2) Python Language for medical image construction and processing.
발표초록 : High-level interpreting languages, such as IDL and matlab, are commonly used for reconstruction and processing of medical images, particularly in the academic community. Although interpreter is generally much slower for a specific process than using a compiling language, including C/C++, they provide superiorities in coding over the compiling language, particularly for visualization. However, these commercial softwares are expensive, particularly for non-academic institute, and the installation is limited to the number of licenses purchased. It is time to switch or start with Python language. Python is an open-source, public domain interpreting language and the number of users are rapidly growing. As is matlab and IDL, the processing is slow, but it can call a C/C++ program for time-consuming process.
In this presentation, I will speak about my experience with image construction and processing using Python. I will introduce two of my research projects, which utilized Python programming: (1). Reconstruction of 3D ultra-short TE (UTE) data and (2). Monte-Carlo simulation of water diffusion in spinal cord.
강연자 성함&직함 / 소속 : Eun-Kee Jeong, Professor of Radiology and Imaging Sciences, Adjunct Professor of Physics and Astronomy, University of Utah, Salt Lake City, UT 84108, USA