ESTRO 2024 - Abstract Book

S4509

Physics - Machine learning models and clinical applications

ESTRO 2024

63.6% of lesions had a complete response and were used as ground truth for the supervised models. LASSO strongly associated complete response with three variables, namely the lesion volume (PTV), the type of lesions (lymph-nodal versus parenchymal), and the biological effective dose (BED10), that were used as input for ML modeling. In the training set, the AUCs for complete response were 0.751 (95% CI: 0.716-0.786), 0.766 (95% CI: 0.729-0.802), and 0.800 (95% CI: 0.742-0.857) for the LR, CART, and SVM with a radial basis function kernel, respectively. These models achieve AUC values of 0.727 (95% CI: 0.669-0.795), 0.734 (95% CI: 0.649-0.815), and 0.771 (95% CI: 0.717-0.824) in the external testing set, demonstrating excellent generalizability. Figure 1 shows the average impact (ie. mean SHAP) of each feature on the model output and the directional SHAP values and relative values of model features. The SHAP analysis indicates that the type of lesion (lymph-nodal vs. parenchymal) had the greatest influence on the model predictions, followed by BED10 and PTV. As shown in Figure 1, lymph-nodal lesions with higher BED10 and smaller PTV are associated with a high chance of complete response. The CART classification tree for the most informative variables is displayed in Figure 2.

Conclusion:

ML models enable a reliable prediction of the treatment response of oligometastatic lesions receiving SBRT. This approach may be able to identify patients who are more likely to have an incomplete response or a higher risk of recurrence and would be of great benefit in customizing treatment plans to the patient’s unique disease profile.