NLM tele-educational application for radiologists to interpret mammography.

Abstract

The goal of this study was to provide unique tools for an educational program to improve the skills, namely consistency and accuracy, of radiology residents who interpret digital mammograms. The tele-educational tools, created at UNC, will be implemented locally and connected to the National Digital Mammography Archive (NDMA) through the Next Generation Internet (NGI). This application includes an annotation tool, as well as a teaching and testing tool. The annotation tool will allow radiologists to label all imaging findings including the specific location information in mammograms, and make lesion diagnosis based on pathology reports. The teaching tool will allow teachers to demonstrate cases of specific types and diagnoses. Trainees themselves will also be able to use the teaching tool for reviewing of cases of types of their choosing. Our testing tool can test radiology residents inverted exclamation mark performance in interpreting digital mammograms, and provides them detailed performance test results, such as, sensitivity, specificity, ROC curves, AUC values, etc. A local Oracle database was designed and implemented at UNC to support those tools. We developed a method to map information of cases from the local database to DICOM Structure Reports by using XML techniques.