Skip to main content
. 2026 Mar 28;16:10811. doi: 10.1038/s41598-026-42492-1

Table 3.

Show the results of MobileNetV2.

Fold Class Precision Recall F1-score ROC-AUC
Fold(1) No Finding 0.69 0.75 0.72 0.86
Pneumonia 0.81 0.87 0.84 0.95
Pneumothorax 0.48 0.81 0.60 0.88
Effusion 0.70 0.78 0.74 0.87
Cardiomegaly 0.60 0.85 0.71 0.94
Fold(1)—validation accuracy 0.88 – – – –
Fold(2) No Finding 0.63 0.81 0.71 0.86
Pneumonia 0.81 0.87 0.84 0.95
Pneumothorax 0.50 0.77 0.61 0.87
Effusion 0.65 0.81 0.73 0.86
Cardiomegaly 0.62 0.86 0.72 0.94
Fold(2)—validation accuracy 0.88 – – – –
Fold(3) No Finding 0.66 0.78 0.71 0.86
Pneumonia 0.79 0.85 0.82 0.95
Pneumothorax 0.52 0.77 0.62 0.88
Effusion 0.69 0.77 0.72 0.87
Cardiomegaly 0.65 0.84 0.74 0.94
Fold(3)—validation accuracy 0.88 – – – –
Class average No Finding 0.66 0.78 0.71 0.87
Pneumonia 0.80 0.87 0.83 0.95
Pneumothorax 0.50 0.78 0.61 0.87
Effusion 0.68 0.79 0.73 0.87
Cardiomegaly 0.63 0.85 0.72 0.94
Class standard deviation No Finding 0.0271 0.0247 0.0055 0.0015
Pneumonia 0.0065 0.0083 0.0073 0.0023
Pneumothorax 0.0170 0.0180 0.0078 0.0031
Effusion 0.0206 0.0205 0.0067 0.0037
Cardiomegaly 0.0213 0.0072 0.0116 0.0019
Average validation accuracy 0.88 – – – –