Construction of mammographic examination process ontology using bottom–up hierarchical task analysis

Abstract

Describing complex mammography examination processes is important for improving the quality of mammograms. It is often difficult for experienced radiologic technologists to explain the process because their techniques depend on their experience and intuition. In our previous study, we analyzed the process using a new bottom-up hierarchical task analysis and identified key components of the process. Leveraging the results of the previous study, the purpose of this study was to construct a mammographic examination process ontology to formally describe the relationships between the process and image evaluation criteria to improve the quality of mammograms. First, we identified and created root classes: task, plan, and clinical image evaluation (CIE). Second, we described an "is-a" relation referring to the result of the previous study and the structure of the CIE. Third, the procedural steps in the ontology were described using the new properties: "isPerformedBefore," "isPerformedAfter," and "isPerformedAfterIfNecessary." Finally, the relationships between tasks and CIEs were described using the "isAffectedBy" property to represent the influence of the process on image quality. In total, there were 219 classes in the ontology. By introducing new properties related to the process flow, a sophisticated mammography examination process could be visualized. In relationships between tasks and CIEs, it became clear that the tasks affecting the evaluation criteria related to positioning were greater in number than those for image quality. We developed a mammographic examination process ontology that makes knowledge explicit for a comprehensive mammography process. Our research will support education and help promote knowledge sharing about mammography examination expertise.

A multi-disciplinary review of knowledge acquisition methods: From human to autonomous eliciting agents

“…Comprehensive reviews of the major body of TD techniques and their applications can be found in [49,190,207,98]. Recently, task diagram methods have been used extensively in various fields, such as in medical domain in clinical research and pathology [209,205], in serious games [10,193], in ergonomics and human factors [149] or in instructional design [214] applications. In the last decade the classic task diagram methods also evolved towards a different elicitation approach called Graphic Elicitation [202] in which diagrams, photos or various forms of visual arts are used for stimulating experts' thoughts on task structure and performance, or in the reverse process experts are asked to provide visual representation of the personal understanding of tasks, concepts, experiences and behaviours.…”

The application of human reliability analysis to carpal tunnel decompression

“…8 , 9 Recent applications of HRA in medicine include adoption by anesthesia, radiology and laparoscopic surgery. 6 , 10 – 13 This study is the first to apply HRA in hand surgery.…”