Medical images are an essential part of diagnostics and treatment planning. The variety of images produced and the amount are rising constantly. Digital radiology has also brought new possibilities for the use of medical images several contexts. In fields such as evidence–based medicine or case–based reasoning medical image data can play a prominent role if tools are available to ease access to images and the accompanying textual data. Retrieval algorithms need to meet the information need of the users at a certain time. The right information needs to be accessible to the right persons at the right time.
The medGIFT project includes several axes around the retrieval of medical images from a variety of databases and image kinds as well as for several applications. The framework is based around the open source image retrieval tool GIFT (GNU Image Finding Tool) and adds tools to this environment to create a system adapted for the domain–specific needs in medical image retrieval. These tools include the preprocessing of images for better retrieval, through the extraction of the main object or even through segmentation in specialised fields such as lung image retrieval. The combination and integration of GIFT with tools for text retrieval such as Lucene and EasyIR are other applications. Another strong point of GIFT is the creation of an infrastructure for image retrieval evaluation. The ImageCLEFmed benchmark is a result of the project and the outcome does not only help locally but is accessible for many research groups on all continents.
No comments:
Post a Comment