Using Google Desktop to Create an Education-centric Search Engine by Integrating MIRC and RIS*

*Award Winning RSNA project Author List: M. Kashif | Abdul Muqeet | Nabile M. Safdar | Khan Mohammad Siddiqu Keywords: MIRC, MIRTH, XHTML, Google desktop, web user interface

	Background One of the limitations in current radiology systems is the inability to collectively search important educational cases. By combining teaching files in local MIRC, patient textual report in RIS and images in PACS, we report on the development of an integrated search engine that can perform keyword based searching in an education-centric manner. This project is an extension to last year’s award-winning informatics exhibit. We used a free open source software, Mirth (http://www.mirthproject.org/) as a middleware, to convert XML-based MIRC teaching files and HL7-based RIS reports to standard XHTML files. Google desktop was used to index XHTML files. Using a scripting language, 1) a function was created to launch the PACS viewer and 2) a web user interface was created to perform searching based on combination of customizable parameters.
	Evaluation
	Our resulting implementation provides an easy way to perform simple keyword based searching both in MIRC and RIS simultaneously. Radiologists have found the search engine extremely useful and a faster way of retrieving important clinical cases in a teaching environment. Although the search engine is proficient in facilitating educational needs, a few limitations were identified, such as data redundancy, manual scheduled updates and inability to perform search around particular context. Implementation of such searching logic requires minimal technical knowledge and can be easily implemented as a fresh installation or as an add-on to previously installed search engine.
	Discussion
	Implementing a combined MIRC and RIS search engine can be achieved by using freely available and open source tools to overcome current limitations of radiology systems. Creating an independent search engine also provides an opportunity to further enable the searching feature in electronic health records.
	Conclusion
	A combined searching system can provide an important tool with which radiologists can retrieve desired educational information from heterogeneous information resources. Such tools can be implemented to search through any HL7 or XML based information system repository.