A New Database for Medical Images and Information

We present a medical image and medical record database for the storage, research, transmission, and evaluation of medical images, as well as tele-medicine applications. Any medical image from a source that supports the DICOM standard can be stored and accessed, as well as associated analysis and annotations. Information and image retrieval can be done based on patient info, date, doctor’s annotations, features in the images, or a spatial combination of features. Secure access and transmission is addressed for telemedicine applications. This database application follows all HIPAA regulations. Introduction Most retrievals in medical image database systems are based on the patient identification information or image modality [1]as it is defined in the DICOM standard [2], and it is hoped that inclusion of other features can improve the effectiveness of this type of system. Archimedes includes retrieval based on features, as well as on patient identification information and image modality. The number of digital medical images is rapidly rising, prompting the need for improved storage and retrieval systems. Image archives and imaging systems are an important economic and clinical factor in the hospital environment [3]. The management and the indexing of these large image and information repositories is becoming increasingly complex. Archimedes is an effort at bringing the latest in information technology to the medical community. Picture Archiving and Communication Systems (PACS) are the main software components used to store and access the large amount of visual data in medical departments. Often, several layer architectures exist for quick short-term access and slow long-term storage [4], and a web-based PACS architecture has been proposed [5]. Web interfaces have been described for medical image databases [6], and Archimedes also includes a web interface. A web interface simplifies the deployment of the system and enables tele-medicine, but requires tighter security measures. The Archimedes system was designed as a web-based system from the start, and provides a platform to evaluate the usefulness and effectiveness of incorporating those changes into PACS. Several frameworks for distributed image management solutions have been developed such as I2Cnet [7, 8]. Image retrieval based on visual features is often proposed but unfortunately little is said about the visual features used or the performance obtained. One competing framework with at least a partial implementation is the IRMA (Image Retrieval in Medical Applications) framework [9, 10]. IRMA enables the classification of images into anatomical regions, modality, and orientation. Some frameworks support telemedicine as well as data and image management [11], which Archimedes supports as well. Archimedes is an image analysis and patient records management tool intended for the use of the medical community. It allows doctors to search for common features in a database of images via innovative combinations of search techniques and algorithms. This system allows the rapid retrieval of images and patient records, and can also find patients with similar images, conditions, or annotations to compare treatment successes. The support of the National Science Foundation under Grants EIA-00-91474 and CCF-0515241, Microsoft Research, and the University of Maryland Graduate Research Board is gratefully acknowledged.