Collection of digital image database in Picture Archiving and Communication System (PACS) has provided a useful means for improved image content accessibility. Idiopathic scoliosis is sometimes found in chest radiography of university students. In order to conduct retrospective studies of such disease, extraction of image features based on image content becomes crucial. To identify images with scoliosis, chest images are retrieved from the PACS image server and processed by an application server for content analysis. Automatic detection of image content can be enhanced by the fuzzy logics for the segmentation of vertebrae on the digital images retrieved from the PACS system. The intelligent classifier then identifies cases of scoliosis automatically. In addition to human interpretation, our method provides a more sensitive and efficient tool for detection of the abnormal feature in chest images.
|Title of host publication||adiological Society of Northern America 90th Scientific Assembly and Annual Meeting|
|Number of pages||1|
|Publication status||Published - 2004|
|Event||adiological Society of Northern America 90th Scientific Assembly and Annual Meeting - Chicago, Illinois, US, New Zealand|
Duration: 27 Nov 2004 → 02 Dec 2004
|Conference||adiological Society of Northern America 90th Scientific Assembly and Annual Meeting|
|Period||27/11/04 → 02/12/04|