You are hereXin Wang

Xin Wang

By admin2019 - Posted on 05 January 2015

Xin Wang, received her PhD degree in Library & Information Science from the University of Missouri (MU), Columbia, MO, in 2012. During the course of pursuing her degrees at MU, she worked as a senior User Experience (UX) researcher at the Information Experience (IE) laboratory for over six years. Currently, Xin Wang works as the Lecturer at the Department of Library & Information Science of University of North Texas. Her research interests include User Experience (UX) Design, Health Informatics, and Data Analytics. Dr. Wang is a committee member of American Society for Information Science and Technology (ASIS&T) and the American Medical Informatics Association (AMIA). Dr. Wang servers as conference paper reviewers for the iConference and International Conference on Society and Information Technologies. Also, she is the journal paper reviewers for the Journal of the American Society for Information Science and Technology (JASIS&T) and the Electronic Library.

Establishing Image Attributes for Designing a Visual Knowledge Infrastructure through Analyzing Medical Image Users’ Classifying Behavior

Due to the upcoming retirement of senior image analysts from the baby boom generation, the imperative need for cultivating new generation professionals is rising (Shyu, Erdelez, & Cho, 2008). How to retain experienced image analysts’ knowledge and allow it to be transferred to successive generations is a significant and challenging issue in the intelligence community (Shyu, et al., 2008). The issue is especially prominent in the community of radiography. Unlike other medical specialties, radiographic technologists’ decisions are heavily relied on visually-based tacit knowledge, untold heuristics, and subtle cues. Therefore, it is critical to develop robust tools that may archive, retrieve, and share various image data (e.g., tomographic image) carrying with domain expert’ tacit knowledge. In order to best serve future-generation image analysts’ knowledge acquisition and exchange among interdisciplinary communities, the ongoing project, VisKM infrastructure, is a content - based image retrieval (CBIR) and knowledge-management system. VisKM is funded by the National Science Foundation (NSF) since 2008 and is carrying on via the collaboration of researchers from three disciplines: computer science, information science, and health informatics.

This present study reports a part of research results from a broader research project. This study focuses on the classifying behavior of medical image users. When users had a vague idea of what images they want to find, there was a need for classification of images based on the abstract concepts. Another challenge for designing information retrieval systems is how to display large groups of records representing documents. Thus, studying user classification behavior can make substantial contribution to the user-centric interface design of image retrieval systems. Examples are designing categories for browsing search, menu categories, and organized result displays.

The purpose of this study is to investigate how do domain experts classify images differently from novices? 40 x-ray images were randomly selected from the UMHC Centricity PACS (Picture Archiving Communication System) and all these x-ray images were uploaded to the online sorting tool. Twenty-seven (27) participants were asked to group these images in order to find these images at a later time. After they were finished with sorting, participants was asked to label a name for each group and provide a brief description of the common characteristics of each group of images. At least two common characteristics (visual cues, semantic judgments, or both) need to be identified in one group. The group labels and the common characteristics were analyzed. As a result, novices and experts preferred image attributes with this sorting task were identified and will be recommended to be integrated to the future design and development of a medical image system.

Shyu, C.-R., Erdelez, S., & Cho, K. (2008).  (pp. 25). University of Missouri National Science Foundation.