Research profile H. (Heejun) Lee

H. (Heejun) Lee

Project: From medical report to visual data stories

Medical text reports form the back-bone of current practice in radiology as a standard means of communicating diagnoses and other findings. Even in the case when they are structured (which is not always so), they contain indirect links to medical images (such as CT scans) which then need to be accessed separately. This unnecessarily slows down the communication process, makes it susceptible to (human) error, and makes it difficult to tailor to different target users (referring physicians, patients).

To improve the effectiveness and efficiency of communication of vital information, we propose to replace, or at least augment, medical text reports with visual data stories seamlessly integrating findings, annotations, measurements, medical images, and potentially also audio tracks, into flexible and interactive visual data stories, or narratives. These visual data stories will help the user focus on the most relevant information, and will enable a better understanding of the data. We will achieve this ambitious goal by combining techniques from human-computer interaction, computer graphics, and (big data) visualisation. Our proof-of-concept implementation of the proposed visual data story tool will be tested and fine-tuned in the context of UMCG and the big medical data there, and is expected to enhance, if not, in the long term, overhaul the current practice based on rigid text reports.

Keywords: medical images, big data visualisation.

Fields of expertise involved: Computer Science, Scientific Visualization and Computer Graphics.