Pulse+IT Magazine - February 2013

Page 48

PULSEITMAGAZINE.COM.AU

048

Feature

SMART SOFTWARE TO AUTOMATE THE

ANALYSIS OF PATHOLOGY AND RADIOLOGY REPORTS Reading and processing narrative-based clinical reports is a time-consuming process. To ease the workload of clinical staff and aid the computer processing of these reports, CSIRO is developing smart clinical decision support software called Medtex. The software is aimed at extracting free text data to aid decision support and take the weight off clinical staff.

ANTHONY NGUYEN BEng (Hons 1), PhD anthony.nguyen@csiro.au GUIDO ZUCCON BEng, MEng, PhD guido.zuccon@csiro.au

Medtex is able to process clinical records such as pathology and radiology reports that contain a lot of valuable medical information that is often buried in text which may be unstructured, ungrammatical or fragmented. An example is determining the stage of a cancer patient. Determining the stage that a cancer is at involves drawing information from a number of sources such as radiology, pathology and surgical reports. This is usually a time-consuming task requiring the expertise of more than one person. A simple way of easily and consistently collecting clinical data could improve health outcomes for patients, boost the efficiency of the health system and provide a rich data set for further research.

About the authors Dr Anthony Nguyen is a Research Team Leader and Dr Guido Zuccon is a Post Doctoral Fellow from the Australian e-Health Research Centre. Anthony and Guido develop medical text analysis capabilities to unlock information in electronic health records for supporting clinical decision making.

Medtex works by “learning� what statements to look for and uses SNOMED CT, the internationally defined set of clinical terms, to unify language across information sources. The more reports the software processes, the smarter it gets.

Virtual cancer registry The medical text analysis software performs cancer registry tasks such as the notification of cancer reports and the coding of notifications data.

The system automatically scans HL7 messages and analyses the free-text reports for terms and concepts relevant to cancer. Classification of pathology reports that are notifiable cancers can be achieved with sensitivities of 98 per cent and specificities of 96 per cent. The coding of specific cancer notification items such as basis of diagnosis, histological type and grade, primary site and laterality can be extracted with overall accuracies of 80 per cent. In the case of lung cancer staging, positive results were achieved after a formal trial on lung cancer cases comparing the stages it assigned with those given by expert pathologists. Medtex also allows for detailed tumour stream synoptic reporting. This software has been developed in conjunction with the Queensland Cancer Control Analysis Team, Queensland Health.

Computer software to read limb X-ray reports The checking of X-ray reports to ensure limb fractures are not missed and that patients receive appropriate follow-up once discharged from the emergency department is essential but can often be a laborious task.