Table 1 Final class-wise distribution of the cleaned Lung X-ray image–clinical text dataset after data preprocessing. Significant values are in [bold].

From: Graph attention network-based multimodal approach for lung diseases classification

Disease class

Disease label

Train

Validation

Test

Total samples

Chest changes

0

6000

891

672

7563

Degenerative infectious diseases

1

6000

948

719

7667

Encapsulated lesions

2

6000

947

693

7640

Higher density

3

6000

960

722

7682

Lower density

4

6000

974

733

7707

Mediastinal changes

5

6000

907

751

7658

Normal

6

6000

1039

719

7758

Obstructive pulmonary diseases

7

6000

953

679

7632

Total

–

48,000

7619

5688

61,307