The importance of planning CT-based imaging features for machine learning-based prediction of pain response
Abstract Patients suffering from painful spinal bone metastases (PSBMs) often undergo palliative radiation therapy (RT), with an efficacy of approximately two thirds of patients. In this exploratory investigation, we assessed the effectiveness of machine learning (ML) models trained on radiomics, se...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-10-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-43768-6 |
| _version_ | 1827711307879546880 |
|---|---|
| author | Óscar Llorián-Salvador Joachim Akhgar Steffi Pigorsch Kai Borm Stefan Münch Denise Bernhardt Burkhard Rost Miguel A. Andrade-Navarro Stephanie E. Combs Jan C. Peeken |
| author_facet | Óscar Llorián-Salvador Joachim Akhgar Steffi Pigorsch Kai Borm Stefan Münch Denise Bernhardt Burkhard Rost Miguel A. Andrade-Navarro Stephanie E. Combs Jan C. Peeken |
| author_sort | Óscar Llorián-Salvador |
| collection | DOAJ |
| description | Abstract Patients suffering from painful spinal bone metastases (PSBMs) often undergo palliative radiation therapy (RT), with an efficacy of approximately two thirds of patients. In this exploratory investigation, we assessed the effectiveness of machine learning (ML) models trained on radiomics, semantic and clinical features to estimate complete pain response. Gross tumour volumes (GTV) and clinical target volumes (CTV) of 261 PSBMs were segmented on planning computed tomography (CT) scans. Radiomics, semantic and clinical features were collected for all patients. Random forest (RFC) and support vector machine (SVM) classifiers were compared using repeated nested cross-validation. The best radiomics classifier was trained on CTV with an area under the receiver-operator curve (AUROC) of 0.62 ± 0.01 (RFC; 95% confidence interval). The semantic model achieved a comparable AUROC of 0.63 ± 0.01 (RFC), significantly below the clinical model (SVM, AUROC: 0.80 ± 0.01); and slightly lower than the spinal instability neoplastic score (SINS; LR, AUROC: 0.65 ± 0.01). A combined model did not improve performance (AUROC: 0,74 ± 0,01). We could demonstrate that radiomics and semantic analyses of planning CTs allowed for limited prediction of therapy response to palliative RT. ML predictions based on established clinical parameters achieved the best results. |
| first_indexed | 2024-03-10T17:52:50Z |
| format | Article |
| id | doaj.art-67160e5a91df44a3901b2e18d26b3925 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-10T17:52:50Z |
| publishDate | 2023-10-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-67160e5a91df44a3901b2e18d26b39252023-11-20T09:18:12ZengNature PortfolioScientific Reports2045-23222023-10-0113111110.1038/s41598-023-43768-6The importance of planning CT-based imaging features for machine learning-based prediction of pain responseÓscar Llorián-Salvador0Joachim Akhgar1Steffi Pigorsch2Kai Borm3Stefan Münch4Denise Bernhardt5Burkhard Rost6Miguel A. Andrade-Navarro7Stephanie E. Combs8Jan C. Peeken9Department of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich (TUM)Department of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich (TUM)Department of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich (TUM)Department of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich (TUM)Department of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich (TUM)Department of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich (TUM)Department for Bioinformatics and Computational Biology, Informatik 12, Technical University of Munich (TUM)Institute of Organismic and Molecular Evolution, Johannes Gutenberg University MainzDepartment of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich (TUM)Department of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich (TUM)Abstract Patients suffering from painful spinal bone metastases (PSBMs) often undergo palliative radiation therapy (RT), with an efficacy of approximately two thirds of patients. In this exploratory investigation, we assessed the effectiveness of machine learning (ML) models trained on radiomics, semantic and clinical features to estimate complete pain response. Gross tumour volumes (GTV) and clinical target volumes (CTV) of 261 PSBMs were segmented on planning computed tomography (CT) scans. Radiomics, semantic and clinical features were collected for all patients. Random forest (RFC) and support vector machine (SVM) classifiers were compared using repeated nested cross-validation. The best radiomics classifier was trained on CTV with an area under the receiver-operator curve (AUROC) of 0.62 ± 0.01 (RFC; 95% confidence interval). The semantic model achieved a comparable AUROC of 0.63 ± 0.01 (RFC), significantly below the clinical model (SVM, AUROC: 0.80 ± 0.01); and slightly lower than the spinal instability neoplastic score (SINS; LR, AUROC: 0.65 ± 0.01). A combined model did not improve performance (AUROC: 0,74 ± 0,01). We could demonstrate that radiomics and semantic analyses of planning CTs allowed for limited prediction of therapy response to palliative RT. ML predictions based on established clinical parameters achieved the best results.https://doi.org/10.1038/s41598-023-43768-6 |
| spellingShingle | Óscar Llorián-Salvador Joachim Akhgar Steffi Pigorsch Kai Borm Stefan Münch Denise Bernhardt Burkhard Rost Miguel A. Andrade-Navarro Stephanie E. Combs Jan C. Peeken The importance of planning CT-based imaging features for machine learning-based prediction of pain response Scientific Reports |
| title | The importance of planning CT-based imaging features for machine learning-based prediction of pain response |
| title_full | The importance of planning CT-based imaging features for machine learning-based prediction of pain response |
| title_fullStr | The importance of planning CT-based imaging features for machine learning-based prediction of pain response |
| title_full_unstemmed | The importance of planning CT-based imaging features for machine learning-based prediction of pain response |
| title_short | The importance of planning CT-based imaging features for machine learning-based prediction of pain response |
| title_sort | importance of planning ct based imaging features for machine learning based prediction of pain response |
| url | https://doi.org/10.1038/s41598-023-43768-6 |
| work_keys_str_mv | AT oscarlloriansalvador theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT joachimakhgar theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT steffipigorsch theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT kaiborm theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT stefanmunch theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT denisebernhardt theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT burkhardrost theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT miguelaandradenavarro theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT stephanieecombs theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT jancpeeken theimportanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT oscarlloriansalvador importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT joachimakhgar importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT steffipigorsch importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT kaiborm importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT stefanmunch importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT denisebernhardt importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT burkhardrost importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT miguelaandradenavarro importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT stephanieecombs importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse AT jancpeeken importanceofplanningctbasedimagingfeaturesformachinelearningbasedpredictionofpainresponse |