Fig. 6: AUC values with 95% confidence intervals across analysis types. | Communications Medicine

Fig. 6: AUC values with 95% confidence intervals across analysis types.

From: Volumetric spline-based Kolmogorov-Arnold architectures surpass CNNs, vision transformers, and graph networks for Parkinson’s disease detection

Fig. 6

a Isolated dataset analysis showing model performance with confidence intervals. b Combined dataset analysis across merged datasets. c Hold-out analysis for cross-dataset generalization. Horizontal bars represent mean AUC values, with blue bars indicating 2D models and orange bars indicating 3D models. n = 59 independent subjects (PPMI), n = 43 (NEUROCON), n = 40 (Tao Wu) for analyses. Error bars represent 95% confidence intervals derived from bootstrap analysis.

Back to article page