An Ensemble Approach to Knowledge-Based Intensity-Modulated Radiation Therapy Planning
Knowledge-based planning (KBP) utilizes experienced planners’ knowledge embedded in prior plans to estimate optimal achievable dose volume histogram (DVH) of new cases. In the regression-based KBP framework, previously planned patients’ anatomical features and DVHs are extracted, and prior knowledge...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2018-03-01
|
| Series: | Frontiers in Oncology |
| Subjects: | |
| Online Access: | http://journal.frontiersin.org/article/10.3389/fonc.2018.00057/full |
| _version_ | 1811264502751035392 |
|---|---|
| author | Jiahan Zhang Q. Jackie Wu Tianyi Xie Yang Sheng Fang-Fang Yin Yaorong Ge |
| author_facet | Jiahan Zhang Q. Jackie Wu Tianyi Xie Yang Sheng Fang-Fang Yin Yaorong Ge |
| author_sort | Jiahan Zhang |
| collection | DOAJ |
| description | Knowledge-based planning (KBP) utilizes experienced planners’ knowledge embedded in prior plans to estimate optimal achievable dose volume histogram (DVH) of new cases. In the regression-based KBP framework, previously planned patients’ anatomical features and DVHs are extracted, and prior knowledge is summarized as the regression coefficients that transform features to organ-at-risk DVH predictions. In our study, we find that in different settings, different regression methods work better. To improve the robustness of KBP models, we propose an ensemble method that combines the strengths of various linear regression models, including stepwise, lasso, elastic net, and ridge regression. In the ensemble approach, we first obtain individual model prediction metadata using in-training-set leave-one-out cross validation. A constrained optimization is subsequently performed to decide individual model weights. The metadata is also used to filter out impactful training set outliers. We evaluate our method on a fresh set of retrospectively retrieved anonymized prostate intensity-modulated radiation therapy (IMRT) cases and head and neck IMRT cases. The proposed approach is more robust against small training set size, wrongly labeled cases, and dosimetric inferior plans, compared with other individual models. In summary, we believe the improved robustness makes the proposed method more suitable for clinical settings than individual models. |
| first_indexed | 2024-04-12T20:05:26Z |
| format | Article |
| id | doaj.art-41080aad4d2d4015bb99d9a127714aa0 |
| institution | Directory Open Access Journal |
| issn | 2234-943X |
| language | English |
| last_indexed | 2024-04-12T20:05:26Z |
| publishDate | 2018-03-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Oncology |
| spelling | doaj.art-41080aad4d2d4015bb99d9a127714aa02022-12-22T03:18:24ZengFrontiers Media S.A.Frontiers in Oncology2234-943X2018-03-01810.3389/fonc.2018.00057329062An Ensemble Approach to Knowledge-Based Intensity-Modulated Radiation Therapy PlanningJiahan Zhang0Q. Jackie Wu1Tianyi Xie2Yang Sheng3Fang-Fang Yin4Yaorong Ge5Department of Radiation Oncology, Duke University Medical Center, Durham, NC, United StatesDepartment of Radiation Oncology, Duke University Medical Center, Durham, NC, United StatesDepartment of Radiation Oncology, Duke University Medical Center, Durham, NC, United StatesDepartment of Radiation Oncology, Duke University Medical Center, Durham, NC, United StatesDepartment of Radiation Oncology, Duke University Medical Center, Durham, NC, United StatesDepartment of Software and Information Systems, University of North Carolina at Charlotte, Charlotte, NC, United StatesKnowledge-based planning (KBP) utilizes experienced planners’ knowledge embedded in prior plans to estimate optimal achievable dose volume histogram (DVH) of new cases. In the regression-based KBP framework, previously planned patients’ anatomical features and DVHs are extracted, and prior knowledge is summarized as the regression coefficients that transform features to organ-at-risk DVH predictions. In our study, we find that in different settings, different regression methods work better. To improve the robustness of KBP models, we propose an ensemble method that combines the strengths of various linear regression models, including stepwise, lasso, elastic net, and ridge regression. In the ensemble approach, we first obtain individual model prediction metadata using in-training-set leave-one-out cross validation. A constrained optimization is subsequently performed to decide individual model weights. The metadata is also used to filter out impactful training set outliers. We evaluate our method on a fresh set of retrospectively retrieved anonymized prostate intensity-modulated radiation therapy (IMRT) cases and head and neck IMRT cases. The proposed approach is more robust against small training set size, wrongly labeled cases, and dosimetric inferior plans, compared with other individual models. In summary, we believe the improved robustness makes the proposed method more suitable for clinical settings than individual models.http://journal.frontiersin.org/article/10.3389/fonc.2018.00057/fulltreatment planningdose volume histogram predictionregression modelmachine learningensemble modelstatistical modeling |
| spellingShingle | Jiahan Zhang Q. Jackie Wu Tianyi Xie Yang Sheng Fang-Fang Yin Yaorong Ge An Ensemble Approach to Knowledge-Based Intensity-Modulated Radiation Therapy Planning Frontiers in Oncology treatment planning dose volume histogram prediction regression model machine learning ensemble model statistical modeling |
| title | An Ensemble Approach to Knowledge-Based Intensity-Modulated Radiation Therapy Planning |
| title_full | An Ensemble Approach to Knowledge-Based Intensity-Modulated Radiation Therapy Planning |
| title_fullStr | An Ensemble Approach to Knowledge-Based Intensity-Modulated Radiation Therapy Planning |
| title_full_unstemmed | An Ensemble Approach to Knowledge-Based Intensity-Modulated Radiation Therapy Planning |
| title_short | An Ensemble Approach to Knowledge-Based Intensity-Modulated Radiation Therapy Planning |
| title_sort | ensemble approach to knowledge based intensity modulated radiation therapy planning |
| topic | treatment planning dose volume histogram prediction regression model machine learning ensemble model statistical modeling |
| url | http://journal.frontiersin.org/article/10.3389/fonc.2018.00057/full |
| work_keys_str_mv | AT jiahanzhang anensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT qjackiewu anensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT tianyixie anensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT yangsheng anensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT fangfangyin anensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT yaorongge anensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT jiahanzhang ensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT qjackiewu ensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT tianyixie ensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT yangsheng ensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT fangfangyin ensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning AT yaorongge ensembleapproachtoknowledgebasedintensitymodulatedradiationtherapyplanning |