Differentiating Gliosarcoma from Glioblastoma: A Novel Approach Using PEACE and XGBoost to Deal with Datasets with Ultra-High Dimensional Confounders.

In this study, we used a recently developed causal methodology, called Probabilistic Easy Variational Causal Effect (PEACE), to distinguish gliosarcoma (GSM) from glioblastoma (GBM). Our approach uses a causal metric which combines Probabilistic Easy Variational Causal Effect (PEACE) with the XGBoost, or eXtreme Gradient Boosting, algorithm. Unlike prior research, which often relied on statistical models to reduce dataset dimensions before causal analysis, our approach uses the complete dataset with PEACE and the XGBoost algorithm. PEACE provides a comprehensive measurement of direct causal effects, applicable to both continuous and discrete variables. Our method provides both positive and negative versions of PEACE together with their averages to calculate the positive and negative causal effects of the radiomic features on the variable representing the type of tumor (GSM or GBM). In our model, PEACE and its variations are equipped with a degree d which varies from 0 to 1 and it reflects the importance of the rarity and frequency of the events. By using PEACE with XGBoost, we achieved a detailed and nuanced understanding of the causal relationships within the dataset features, facilitating accurate differentiation between GSM and GBM. To assess the XGBoost model, we used cross-validation and obtained a mean accuracy of 83% and an average model MSE of 0.130. This performance is notable given the high number of columns and low number of rows (code on GitHub).

Reader reaction to "Outcome-adaptive lasso: Variable selection for causal inference" by Shortreed and Ertefaie (2017).

Shortreed and Ertefaie introduced a clever propensity score variable selection approach for estimating average causal effects, namely, the outcome adaptive lasso (OAL). OAL aims to select desirable covariates, confounders, and predictors of outcome, to build an unbiased and statistically efficient propensity score estimator. Due to its design, a potential limitation of OAL is how it handles the collinearity problem, which is often encountered in high-dimensional data. As seen in Shortreed and Ertefaie, OAL's performance degraded with increased correlation between covariates. In this note, we propose the generalized OAL (GOAL) that combines the strengths of the adaptively weighted L1 penalty and the elastic net to better handle the selection of correlated covariates. Two different versions of GOAL, which differ in their procedure (algorithm), are proposed. We compared OAL and GOAL in simulation scenarios that mimic those examined by Shortreed and Ertefaie. Although all approaches performed equivalently with independent covariates, we found that both GOAL versions were more performant than OAL in low and high dimensions with correlated covariates.