Click here to


Are you sure ?

Yes, do it No, cancel

Assessment of Plan Quality Variability in Head and Neck Treatments Using Feasibility Predictions in a Commercial Software

S Ahmed*, C Liu, D LaHurd, E Murray, M Kolar, N Joshi, N Woody, S Koyfman, P Xia, The Cleveland Clinic Foundation, Cleveland, OH


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

Room: AAPM ePoster Library

Purpose: reduce variability in the plan quality for Head and Neck (HN) treatments among different centers of our department using a commercial feasibility prediction tool.

Methods: total of 109 HN plans treated between 2017-2019 were selected for this study, with 52 plans from the main center and 57 from regional centers. The DICOM data (CT images, RT-Structures and RT dose) were imported into the commercial software. For each plan, a benchmark dose was calculated with predicted DVHs for OARs at various feasibility levels (f-value) with lower f-values indicating increasing difficulty to achieve. DVHs from clinical plans for ten OARs were compared to the feasibility DVHs with f-value ranging from 0.0-0.5. f-values were sampled for maximum Dice coefficient. Using an auto-planning tool, five HN cases were re-planned using feasibility DVHs as inputs for dose objectives in the optimizer. The Wilcoxon rank sum test was used for statistical analysis.

Results: the entire sample, the median f-values were 0.28 and 0.33 for main and regional plans respectively. The median f-values were statistically different for all OARs except left parotid and mandible. The OARs with noticeable differences in median f-values, between main and regional centers, were spinal cord (0.25), esophagus (0.19), larynx (0.15), supraglottis (0.14) and trachea (0.07), indicating potential for further OAR sparing. In five re-planned cases, the overall median f-value decreased from 0.39 to 0.23 with similar dose coverage to the tumor volumes. Large differences between the f-values of re-plans and clinical plans were observed for spinal cord (0.01-0.41), esophagus (-0.03-0.29), larynx (0-0.3), supraglottis (0.06-0.25) and trachea (0.06-0.33). This translated in the average reduction of 8.57Gy, 8.66Gy, 8.43Gy, 7.36Gy and 7.45Gy respectively.

Conclusion: DVH was found a practicable tool for the assessment of plan quality and could serve as guidance for reducing the plan quality variability for HN treatments.

Funding Support, Disclosures, and Conflict of Interest: PX received a research grant and one travel support from Philips Radiation Oncology Systems, Fitchburg, WI. SA was partially supported by the research grant from Philips. All other authors:CWL, DL, EM, MK,NJ,NW and SK have no funding disclosures and conflict of interest.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email