Machine Learning Radiomics Model for External and Internal Respiratory Motion Correlation Prediction in Lung Tumor
Objectives: The complexity and specificity of lung tumor motion render it necessary to determine the external and internal correlation individually before applying indirect tumor tracking. However, the correlation cannot be determined from patient respiratory and tumor clinical characteristics befor...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SAGE Publishing
2022-12-01
|
| Series: | Technology in Cancer Research & Treatment |
| Online Access: | https://doi.org/10.1177/15330338221143224 |
| _version_ | 1811300722223874048 |
|---|---|
| author | Xiangyu Zhang MM Xinyu Song MM Guangjun Li MS Lian Duan BE Guangyu Wang MM Guyu Dai MM Ying Song PhD Jing Li PhD Sen Bai MS |
| author_facet | Xiangyu Zhang MM Xinyu Song MM Guangjun Li MS Lian Duan BE Guangyu Wang MM Guyu Dai MM Ying Song PhD Jing Li PhD Sen Bai MS |
| author_sort | Xiangyu Zhang MM |
| collection | DOAJ |
| description | Objectives: The complexity and specificity of lung tumor motion render it necessary to determine the external and internal correlation individually before applying indirect tumor tracking. However, the correlation cannot be determined from patient respiratory and tumor clinical characteristics before treatment. The purpose of this study is to present a machine learning model for an external/internal correlation prediction that is based on computed tomography (CT) radiomic features. Methods: 4-dimensional computed tomography (4DCT) images of 67 patients were collected retrospectively, and the external/internal correlation of lung tumor was calculated based on Spearman's rank correlation coefficient. Radiomic features were extracted from average intensity projection and the light gradient boosting machine (LightGBM)-based cross-validation (the recursive elimination method) was used for feature selection. The LightGBM framework forecasting models with classification thresholds 0.7, 0.8, and 0.9 are established using stratified 5-fold cross-validation. Model performance was assessed using receiver operating characteristics, sensitivity, and specificity. Results: There were 16, 18, and 13 features selected for models 0.7, 0.8, and 0.9, respectively. Texture features are of great importance in external/internal correlation prediction compared to other features in all models. The sensitivities of the predictions in models 0.7, 0.8, and 0.9 were 0.800 ± 0.126, 0.829 ± 0.140, and 0.864 ± 0.086, respectively. The specificities were 0.771 ± 0.114, 0.936 ± 0.0581, and 0.839 ± 0.101, whereas the area under the curve (AUC) was 0.837, 0.946, and 0.877, respectively. Conclusions: Our findings indicate that radiomics is an effective tool for respiratory motion correlation prediction, which can extract tumor motion characteristics. We proposed a machine learning framework for correlation prediction in the motion management strategy for lung tumor patients. |
| first_indexed | 2024-04-13T06:55:38Z |
| format | Article |
| id | doaj.art-04c24c7b726947418044a14d003c2415 |
| institution | Directory Open Access Journal |
| issn | 1533-0338 |
| language | English |
| last_indexed | 2024-04-13T06:55:38Z |
| publishDate | 2022-12-01 |
| publisher | SAGE Publishing |
| record_format | Article |
| series | Technology in Cancer Research & Treatment |
| spelling | doaj.art-04c24c7b726947418044a14d003c24152022-12-22T02:57:15ZengSAGE PublishingTechnology in Cancer Research & Treatment1533-03382022-12-012110.1177/15330338221143224Machine Learning Radiomics Model for External and Internal Respiratory Motion Correlation Prediction in Lung TumorXiangyu Zhang MM0Xinyu Song MM1Guangjun Li MS2Lian Duan BE3Guangyu Wang MM4Guyu Dai MM5Ying Song PhD6Jing Li PhD7Sen Bai MS8 Radiotherapy Physics and Technology Center, Cancer Center, , Chengdu, China Department of Radiation Oncology, Cancer Center, The , Guangzhou, China Radiotherapy Physics and Technology Center, Cancer Center, , Chengdu, China Department of Radiation Oncology, Perelman School of Medicine, , Philadelphia, PA, USA Radiotherapy Physics and Technology Center, Cancer Center, , Chengdu, China Radiotherapy Physics and Technology Center, Cancer Center, , Chengdu, China Radiotherapy Physics and Technology Center, Cancer Center, , Chengdu, China Radiotherapy Physics and Technology Center, Cancer Center, , Chengdu, China Department of Radiation Oncology, Cancer Center, , Chengdu, ChinaObjectives: The complexity and specificity of lung tumor motion render it necessary to determine the external and internal correlation individually before applying indirect tumor tracking. However, the correlation cannot be determined from patient respiratory and tumor clinical characteristics before treatment. The purpose of this study is to present a machine learning model for an external/internal correlation prediction that is based on computed tomography (CT) radiomic features. Methods: 4-dimensional computed tomography (4DCT) images of 67 patients were collected retrospectively, and the external/internal correlation of lung tumor was calculated based on Spearman's rank correlation coefficient. Radiomic features were extracted from average intensity projection and the light gradient boosting machine (LightGBM)-based cross-validation (the recursive elimination method) was used for feature selection. The LightGBM framework forecasting models with classification thresholds 0.7, 0.8, and 0.9 are established using stratified 5-fold cross-validation. Model performance was assessed using receiver operating characteristics, sensitivity, and specificity. Results: There were 16, 18, and 13 features selected for models 0.7, 0.8, and 0.9, respectively. Texture features are of great importance in external/internal correlation prediction compared to other features in all models. The sensitivities of the predictions in models 0.7, 0.8, and 0.9 were 0.800 ± 0.126, 0.829 ± 0.140, and 0.864 ± 0.086, respectively. The specificities were 0.771 ± 0.114, 0.936 ± 0.0581, and 0.839 ± 0.101, whereas the area under the curve (AUC) was 0.837, 0.946, and 0.877, respectively. Conclusions: Our findings indicate that radiomics is an effective tool for respiratory motion correlation prediction, which can extract tumor motion characteristics. We proposed a machine learning framework for correlation prediction in the motion management strategy for lung tumor patients.https://doi.org/10.1177/15330338221143224 |
| spellingShingle | Xiangyu Zhang MM Xinyu Song MM Guangjun Li MS Lian Duan BE Guangyu Wang MM Guyu Dai MM Ying Song PhD Jing Li PhD Sen Bai MS Machine Learning Radiomics Model for External and Internal Respiratory Motion Correlation Prediction in Lung Tumor Technology in Cancer Research & Treatment |
| title | Machine Learning Radiomics Model for External and Internal Respiratory Motion Correlation Prediction in Lung Tumor |
| title_full | Machine Learning Radiomics Model for External and Internal Respiratory Motion Correlation Prediction in Lung Tumor |
| title_fullStr | Machine Learning Radiomics Model for External and Internal Respiratory Motion Correlation Prediction in Lung Tumor |
| title_full_unstemmed | Machine Learning Radiomics Model for External and Internal Respiratory Motion Correlation Prediction in Lung Tumor |
| title_short | Machine Learning Radiomics Model for External and Internal Respiratory Motion Correlation Prediction in Lung Tumor |
| title_sort | machine learning radiomics model for external and internal respiratory motion correlation prediction in lung tumor |
| url | https://doi.org/10.1177/15330338221143224 |
| work_keys_str_mv | AT xiangyuzhangmm machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor AT xinyusongmm machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor AT guangjunlims machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor AT lianduanbe machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor AT guangyuwangmm machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor AT guyudaimm machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor AT yingsongphd machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor AT jingliphd machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor AT senbaims machinelearningradiomicsmodelforexternalandinternalrespiratorymotioncorrelationpredictioninlungtumor |