Background Tumor heterogeneity is a significant factor contributing to the marked differences in survival rates among glioblastoma multiforme (GBM) patients, who face a poor prognosis. To improve personalized treatment, it is essential to identify specific tumor characteristics that capture this variability and aid in predicting survival. This study aimed to evaluate the utility of dosiomics and radiomics in predicting overall survival (OS). The central hypothesis was that integrating dosiomics and radiomics could improve survival outcome predictions. Methods A total of 74 GBM patients from The Cancer Imaging Archive were retrospectively included. Dosiomic features from the gross tumor volume (GTV) of planned dose distributions, along with radiomic features from the contrast-enhanced tumor (CET) and edema/non-contrast-enhanced tumor (ED/nCET) subregions across various pre-radiation MRI modalities, were extracted and optimized using L1-based feature selection. Logistic Regression (LR) models were built utilizing different feature configurations to assess the discriminative power of dosiomic and radiomic features, considering the impact of heterogeneous subregions. Model performance was assessed through stratified 10-fold cross-validation (CV). Results The dosiomic model exhibited a mean area under the receiver operating characteristic (ROC) curve (AUC) of 0.80 <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\( \pm \)</EquationSource> </InlineEquation> 0.12. The subregion-based models demonstrated mean AUC values of 0.90 <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\( \pm \)</EquationSource> </InlineEquation> 0.09 for the CET subregion and 0.76 <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\( \pm \)</EquationSource> </InlineEquation> 0.10 for the ED/nCET subregion, indicating that the CET subregion significantly outperformed the ED/nCET subregion (p-value < 0.05). The mean AUC values for modality-based models were as follows: 0.86 <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\( \pm \)</EquationSource> </InlineEquation> 0.12 for T1-weighted contrast-enhanced (T1CE), 0.84 <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\( \pm \)</EquationSource> </InlineEquation> 0.18 for T1-weighted (T1), 0.85 <InlineEquation ID="IEq6"> <EquationSource Format="TEX">\( \pm \)</EquationSource> </InlineEquation> 0.14 for T2-weighted (T2), and 0.76 <InlineEquation ID="IEq7"> <EquationSource Format="TEX">\( \pm \)</EquationSource> </InlineEquation> 0.21 for fluid-attenuated inversion recovery (FLAIR) sequences. There was no significant difference in discrimination power among the four modalities (p-value > 0.05). The combined dosiomic and CET model improved performance to 0.96 <InlineEquation ID="IEq8"> <EquationSource Format="TEX">\( \pm \)</EquationSource> </InlineEquation> 0.07 (p < 0.05). Conclusions Dosiomic and pre-radiotherapy MRI-derived radiomic features are capable of stratifying GBM patients into two long-term and short-term groups. Notably, the integration of dosiomics and radiomics significantly enhances survival prediction in GBM patients.

Machine learning-based integration of dosiomics and pre-radiotherapy multimodal MRI radiomics for survival stratification in patients with glioblastoma multiforme

期刊论文
出版模式：

开放获取
发表日期：

2025年12月2日

查看全文

Atefeh Mahmoudi,
Arash Zare Sadeghi,
Hamed Iraji,
Maedeh Barahman,
Pegah Saadatmand,
Elmira Yazdani,
Seied Rabi Mahdavi

摘要

Background

Tumor heterogeneity is a significant factor contributing to the marked differences in survival rates among glioblastoma multiforme (GBM) patients, who face a poor prognosis. To improve personalized treatment, it is essential to identify specific tumor characteristics that capture this variability and aid in predicting survival. This study aimed to evaluate the utility of dosiomics and radiomics in predicting overall survival (OS). The central hypothesis was that integrating dosiomics and radiomics could improve survival outcome predictions.

Methods

A total of 74 GBM patients from The Cancer Imaging Archive were retrospectively included. Dosiomic features from the gross tumor volume (GTV) of planned dose distributions, along with radiomic features from the contrast-enhanced tumor (CET) and edema/non-contrast-enhanced tumor (ED/nCET) subregions across various pre-radiation MRI modalities, were extracted and optimized using L1-based feature selection. Logistic Regression (LR) models were built utilizing different feature configurations to assess the discriminative power of dosiomic and radiomic features, considering the impact of heterogeneous subregions. Model performance was assessed through stratified 10-fold cross-validation (CV).

Results

The dosiomic model exhibited a mean area under the receiver operating characteristic (ROC) curve (AUC) of 0.80 \( \pm \) 0.12. The subregion-based models demonstrated mean AUC values of 0.90 \( \pm \) 0.09 for the CET subregion and 0.76 \( \pm \) 0.10 for the ED/nCET subregion, indicating that the CET subregion significantly outperformed the ED/nCET subregion (p-value < 0.05). The mean AUC values for modality-based models were as follows: 0.86 \( \pm \) 0.12 for T1-weighted contrast-enhanced (T1CE), 0.84 \( \pm \) 0.18 for T1-weighted (T1), 0.85 \( \pm \) 0.14 for T2-weighted (T2), and 0.76 \( \pm \) 0.21 for fluid-attenuated inversion recovery (FLAIR) sequences. There was no significant difference in discrimination power among the four modalities (p-value > 0.05). The combined dosiomic and CET model improved performance to 0.96 \( \pm \) 0.07 (p < 0.05).

Conclusions

Dosiomic and pre-radiotherapy MRI-derived radiomic features are capable of stratifying GBM patients into two long-term and short-term groups. Notably, the integration of dosiomics and radiomics significantly enhances survival prediction in GBM patients.