Click here to


Are you sure ?

Yes, do it No, cancel

Learning From the Past: Features and Functionality of a Software Designed to Search Peer-Reviewed Cases of Treatments Plans

E Schreibmann*, M Abugideiri, T Liu, N Esiashvili, Department of Radiation Oncology and Winship Cancer Institute of Emory University, Atlanta, GA


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: develop a customized software (search engine tool) to browse images, contours, and dose distributions of patients from the clinical database as a tool to aid physicians in the segmentation and treatment planning process by driving inspiration from approaches that worked on similar cases.

Methods: The search includes a database of 12074 peer-reviewed cases treated in the past 3 years in our institution. The search terms are created by analyzing the prescription into structured data of 16 searchable indicators that include patient information, treatment history, and clinical management. This information is presented in customized software allowing clinicians to search by keywords, custom-filter results and review plan details such as segmentation, arcs placements placement, and optimization constraints. Plan visualization is designed for minimizing review times with features such as automatically highlighting met/unmet constraints, automated bookmarks of the planning target volume (PTV) and organ-at-risk (OAR) overlaps or fast browsing of image sets. Technically, the software uses C++ and VTK to visualize most aspects of a treatment plan while clinical data is extracted from Eclipse using scripting.

Results: The software has been used 73 by physics and physician residents across five campuses since November 2019 receiving positive feedback. The residents reported a wide range of applications from designing treatment plans for new patients based on previous experience to research-related tasks such as creating a list of patients that were managed in a specific criterion. Most importantly, it is used to seek similar cases as a guidance tool while managing uncommon cases.

Conclusion: The proposed search engine was found to be a useful learning tool to search previously treated cases. Employing advanced visualization modules for plan review in clinical practice, this search tool allows physicians to learn from past cases, particularly for managing difficult/rare diseases.


Not Applicable / None Entered.


IM/TH- Formal Quality Management Tools: General (most aspects)

Contact Email