Picture archiving and communication systems (PACS) are designed to provide the radiologist with image information he needs. Currently, state-of-art standard of digital imaging and communication in medicine (DICOM) gives only alphanumerical descriptions of image and this is the only information used in PACS to select relevant images. However, it is industry standard now, textual descriptions are insufficient to describe variety of details in medical image. The content-based image retrieval (CBIR) systems could support DICOM-based retrieval systems and fulfilI "right time, right place" paradigms. In our work we developed Image Shark - a content-based image retrieval system, integrated with PACS solution in one, fully-integrated environment, with novel JPEG2000 codec for effective image transmission and interactive communication. Our system joins cIassic PACS with content-based image retrieval engine. Such approach gives very flexible and effective image retrieval ways. A set of image- and wavelet-domain based indexes were implemented, verified in initial experiments and selected as suitable for mammograms, radiograms and other modalities. Precision of data retrieval was comparable with other engines (IRMA, others), giving very promising results with similar or better precision level.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.