29 August 2025 : Database Analysis
Multimodal Neural Assessment for Predicting Post-Stroke Dysphagia: Identifying Critical Risk Factors Using Combined Indicators
Qian Zhang ACEF 1,2, Shuang Wu A 1,2*, Yangmei Shi BCD 2, Qian Chen BCD 2, Jiajie Gao BCD 2, Sijie Sun B 1, Jing Zhang B 1DOI: 10.12659/MSM.948637
Med Sci Monit 2025; 31:e948637
Table 5 ROC evaluation of the predictive value of risk factors for PSD.
| Variables | AUC | P value | 95% CI | Best cut-off value | Sensitivity | Specificity | Youden index |
|---|---|---|---|---|---|---|---|
| NIHSS | 0.663 | <0.001 | 0.599~0.727 | 9.500 | 0.403 | 0.859 | 0.262 |
| Cp | 0.590 | 0.012 | 0.516~0.663 | 0.625 | 0.905 | 0.354 | 0.259 |
| PA | 0.709 | <0.001 | 0.648~0.770 | 79.700 | 0.552 | 0.818 | 0.370 |
| ipsilateral MEP | 0.872 | <0.001 | 0.833~0.911 | 141.635 | 0.736 | 0.889 | 0.625 |
| contralateral MEP | 0.651 | <0.001 | 0.585~0.717 | 170.835 | 0.841 | 0.414 | 0.255 |
| Cp+PA+ipsilateral MEP +contralateral MEP | 0.985 | <0.001 | 0.974~0.995 | 0.737 | 0.920 | 0.960 | 0.880 |
| Cp+PA+ipsilateral MEP | 0.939 | <0.001 | 0.915~0.964 | 0.759 | 0.806 | 0.919 | 0.725 |
| Cp+PA+contralateral MEP | 0.794 | <0.001 | 0.740~0.848 | 0.679 | 0.721 | 0.768 | 0.489 |
| Cp+PA | 0.744 | <0.001 | 0.683~0.846 | 0.556 | 0.866 | 0.566 | 0.432 |
| Cp+ipsilateral MEP | 0.883 | <0.001 | 0.846~0.921 | 0.717 | 0.776 | 0.869 | 0.645 |
| Cp+contralateral MEP | 0.679 | <0.001 | 0.615~0.743 | 0.751 | 0.443 | 0.828 | 0.271 |
| PA+ipsilateral MEP | 0.934 | <0.001 | 0.908~0.959 | 0.855 | 0.721 | 0.980 | 0.701 |
| PA+contralateral MEP | 0.766 | <0.001 | 0.711~0.821 | 0.789 | 0.498 | 0.889 | 0.387 |
| NIHSS – National Institute of Health Stroke Scale; Cp – clustering coefficient; PA – peak amplitude; MEP – motor-evoked potentials. | |||||||






