Additional file 7 of Machine learning-based prediction of short-term outcomes in aneurysmal subarachnoid hemorrhage: a multicenter study integrating clinical and inflammatory indicators

Additional file 7. Fig S6: LIME interpretation of the GBM model. (A) Ranked feature contributions: bar length indicates contribution magnitude, and colors denote direction of effect. The top nine predictors were CLR, WFNS, PNI, GCS, PLR, modified Fisher grade, procalcitonin, NAR, and SII. (B) Example case prediction: probability of favorable outcome 0.89 and unfavorable outcome 0.11. WFNS ≤ 1 and higher GCS contributed positively, whereas elevated CLR had a negative impact.Note:PNI 1 (48.8), NAR 1 (0.31), PLR 1 (295),SII 1 (3037.3),SIRI 1 (9.68), Procalcitonin 1 (<0.04),2