Background
The method of solely using a black box model for radiation toxicity prediction in patients with lung cancer has limitations in explaining the causality of the prediction results. Therefore, the feature importance of predictors was analyzed using explainable artificial intelligence.
Materials and Methods
Predictions were made for the clinical prognosis through SHAP analysis (Shapley additive explanations) by using pneumonia, interstitial lung disease, chronic obstructive pulmonary disease, concurrent chemoradiation therapy, age, and dosimetric factors [lung volume receiving ≥20 Gy (V20), mean lung dose (MLD)] as prognostic factors in 110 lung cancer patients who received radiation therapy. The model was analyzed using a random forest regressor and a tree explainer; and the SHAP analysis was used to examine the features of prognostic factors affecting radiation side effects and to derive mutual impact.
Results
For patients with grade 2 toxicity, pneumonia, MLD, and V30 were analyzed as very important factors in the prediction results. However, for grade 0 toxicity patients, V30 and MLD were identified as the predictors that had a more important impact (SHAP value=0.7) than pneumonia. In addition, pneumonia had a decisive influence on the prognosis for future side effects of grades 1 and 2 or higher, and MLD was found to have a correlation with pneumonia and SHAP value=0.38. Moreover, through this prediction model, the predicted result for patients with mild radiation pneumonitis by the ground truth of a specific patient was 1.6 (close to grade2), and MLD>20 Gy, pathology, V20, and V30 were analyzed as high-risk factors in predicting radiation side effects. The accuracy, sensitivity and specificity of the model system were 0.88, 0.79, and 0.78, respectively.
Conclusion
Through this study, MLD and V30 were analyzed as important predictors of side effects, and the features of each factor were analyzed for the degree of importance by the SHAP value. To predict radiation pneumonia using this method, a personalized analysis was conducted to identify the factors that influenced each patient. Through this process, comparisons were made with the existing black box method, which confirmed that increasing the explainability can reinforce an in-depth analysis of radiation side effect prediction.