Medical imaging technology has experienced radical development in recent two decades, which enables the generation of multiple, large, multi-dimensional medical data. Though those new features promote the clinical applications, they are not well supported by previous DICOM information object definitions (IODs). Consequently, DICOM has provided various kinds of new enhanced information objects (EIOs) since 2001, including enhanced MR, MR Spectroscopy, Enhanced CT, Enhanced XA, Enhanced XRF, X-ray 3D Angiographic, X-Ray 3D Craniofacial, Ophthalmic Tomography, and Enhanced PET. Comparing to the previous information objects, those EIOs have several characteristics: (1) Large amounts of new mandatory attributes are defined for reducing the dependence on private attributes, and for allowing advanced applications to interoperate between different vendors. Most attributes are grouped into hierarchies, which serves for avoiding the redundancy and keeping diversity among frames.(2) Since the increasing complexity of spatial, temporal, and other relationships between frames, dimension information is defined for explicitly showing their dimension organization. (3) To meet clinical requirements, non-image data, such as MR Spectroscopy, is standardized for transfer, display and analysis. (4) New multi-frame mechanism brings many potential technical benefits such as faster network transfers, reduced loading time, and less storage etc. Although all of those characteristics of EIOs are appealing, they still remain on the paper to us. And unfortunately, we can not find any image viewing toolkits that support the EIOs at the time we started. In order to facilitate the recognition of EIOs, we first designed and implemented a viewer for EIOs, named EDViewer. Then we evaluated some of the important features of EIOs based on this application. Methods According to the new features of EIOs described above, we first analyze the requirements carefully and decide the functions of EDViewer as following: (1) Parse the header of new multi-frame images and exhibit all the DICOM information with obvious hierarchy (e.g. tree structure) (2) Display according to the context information that includes image types, functional groups, and dimensions. (3) Support the display of non-image data, as well as images. (4) Transmit multiframe medical data over network. We think these functions can reflect the most typical features of EIOs. The architecture of EDViewer is illustrated in Fig. 1. We adopted a component-based approach. Two ActiveX controls are utilized: MIViewer is a component providing basic image displaying and processing functions for DICOM images; and DisCLdcm is a component supporting various DICOM services, include verification, storage, query/retrieve, and modality worklist. All of these functions can be used for EDViewer directly. In addition, according to our analysis on EIOs, we found that the definitions are based on a common multi-frame mechanism which combines concatenation, dimension organization and so on. Thus, we developed a share...