ActiveDICOM - enhancing static medical images with interaction

Abstract

Digital Imaging and Communications in Medicine (DICOM) is a well-establish standard in medical imaging, consisting not only of image data, but sensitive data such as patient and examination information. Although having a large variety of advanced rendering techniques available, DICOM images are nowadays still generated and sent to the Picture Archiving and Communication System (PACS). These images are then fetched by the medical doctor from a workstation and used for medical reporting. The user has no other possibilities than being able to change the windowing function for displaying the DICOM images. If a certain region is of special interest, either images of the whole data set are generated or have to be specifically requested. Both approaches consume a considerable amount of time. First, one has to inspect the whole data and has to constantly focus on a particular region without being distracted. Secondly, the image generation on demand remains pending until done by the responsible assistant. Despite supporting a broad range of features and being widely applied, DICOM images remain static. Currently, and to the best of our knowledge, there are no means to enhance these static images with interactive capabilities. Providing the opportunity to interact with an image would, inevitably, enhance the medical reporting procedure. In this thesis, we propose a visualization mapping language, called Active DICOM Script (ADICT), which enhances conventional DICOM images with interactive elements. This language acts as an abstraction layer to provide a transparent interface between heterogeneous data, interaction, and visualization. Since these images are becoming active, we refer to them as Active Digital Imaging and Communications inMedicine (ActiveDICOM). Within this thesis we describe how interactions will be encoded into the image while retaining compatibility to common DICOM viewers. Furthermore with the help of ActiveDICOM we can even introduce Visual Analytics to DICOM. We demonstrate our interaction-to-visualizationmapping on a client-server web-based viewer, as this would aid integration within the clinical environment and on mobile devices such as tablets or smart phones.