Content Based Retrieval of Images with Consolidation from Chest X-Ray Databases
There are large amounts of digitized radiographs available with related patient pathology and medical history. Retrieval of archived images are useful for aiding diagnosis and to provide relevant evidence from previous cases, as well as a training mechanism for junior radiologists. Most diseases an...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2015
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | There are large amounts of digitized radiographs available with related patient pathology and medical history. Retrieval of archived images are useful for aiding diagnosis and to provide relevant evidence from previous cases, as well as a training mechanism for junior radiologists. Most diseases and abnormalities tend to appear at specific regions of the image; hence a retrieval system with local features becomes necessary. Medical image segmentation also plays an important role by automating the delineation of anatomical structures. Thus, the main motivation of this thesis is to develop a fully automated lung segmentation approach together with consolidation detection and classification, to be used in a content-based medical image retrieval (CBMIR) system to identify infection and fluid regions in CXR images. Developing a fully automated segmentation approach for a CBMIR system is a challenging task as chest radiography images from different machines produce different contrast and intensity levels, and are also subject to different patient positioning and image projection. |
|---|