S Li - research works

The five traditional ML models showed varying performance, with RF demonstrating the highest average AUC in the training set (AUC = 0.795, 95% CI 0.709-0.881).

GGT_3M, ALB_3M, DBIL_3M, DBIL_1M and GLO_1W were identified as the most critical predictive features.

A TL-based clinical decision support system was deployed, providing dynamic visualization of individualized 24-month NLS trajectories.

Ablation studies confirmed the critical contributions of the CBAM and ASPP modules.

Our model significantly outperformed baseline and state‑of‑the‑art architectures across major structures (P < 0.01, Bonferroni‑corrected), achieving a 17.0% relative improvement in mIoU for the ligamentum flavum compared to DeepLabV3+.

Sensitivity analyses confirmed model robustness, and subgroup analyses revealed consistent performance across major spinal regions.

98% (96-98%); Hodges-Lehmann difference 1.0%, 95% CI: 0.0-2.0%; P=0.02].

Arterial partial pressure of carbon dioxide (PaCO 2 ) during sedation did not differ significantly between groups.

The incidence of hypotension, vasoactive drug use, airway interventions, postoperative organ injury markers, oxygen-interface tolerance scores, intensive care unit (ICU) admission, and neonatal outcomes were similar.

Higher urinary calprotectin levels were observed in the severe AKI group at any time point (All P<0.001).

The AUC of the clinical model increased from 0.693 (95%CI, 0.633 to 0.754) to 0.730 (95%CI, 0.677 to 0.784) upon adding urinary calprotectin, the NRI was 0.337 (P<0.001) and the IDI was 0.055 (P<0.001).