Standardisation in medical image management

Abstract

A medical image without the relevant associated data is of no value. What is relevant depends on the use which is to be made of the image. The associated data may be divided into groups associated with the patient, the image data (array size, data type), the image itself (e.g. acquisition method), the requested imaging procedure to which it belongs, graphical, numerical and time information associated with the images, relevant results of other medical procedures, etc. Images, digital images and the associated data must be managed in an efficient manner in order to deliver appropriate patient care. In order to realise fully the potential of digital imaging, digital image data must be transmitted from acquisition units and stored in such a way as to allow appropriate access to health care practitioners in the imaging department, hospital and community. Standards exist for the formatting and transmission of image and related data (Digital Imaging and Communications in Medicine (DICOM). NEMA PS 3.1-PS3.12, The National Electrical Manufacturers Association, Rosslyn, VA, 1992, 1993, 1995; Health Level Seven, An Application Protocol for Electronic Data Exchange In Healthcare Environment, version 2.2, Health Level Seven Inc., Ann Arbor, MI, 1994; Comité European de Normalisation, Technical Committee 251: Medical Informatics) although further work is required in order to improve interoperability. There is some overlap in current standards, thus an important area is the agreement on the scope of the work to be done by different bodies. Standards are lacking in the areas of image management and presentation. Strategies for the efficient use of a memory hierarchy to achieve satisfactory archiving and access must be developed and this depends on the existence of appropriate descriptive data fields. Specific issues exist regarding security of image data. Key areas requiring immediate work are the profiling of descriptive data elements according to different clinical application needs in order to improve interoperability and the development of strategies for evolution from legacy systems to modern networked systems. The best way forward is to consolidate and refine existing specifications.