Fig. 7

a The calibration curves of the AE-CTD-ILD diagnostic nomogram. The “Ideal line” represents the optimal predictive performance, while the “Apparent” represents the prediction probability curve fitted by the nomogram model. The “Bias correction” refers to the prediction probability curve after adjusting for overfitting. The “Apparent” represents the prediction probability curve fitted by the nomogram model, and the “Bias correction” represents the prediction probability curve after correcting for overfitting, thereby indicating the predictive performance of the nomogram model. b The decision curve of the nomogram model, serving as a tool to evaluate the practicality and benefits of the model. The decision curve analysis indicates that the developed nomogram in this study exhibits substantial benefits, suggesting a viable potential application value for clinical decision-making