Skip to main content
European Radiology Experimental

Table 3 Breakdown of labels used for multilabel diagnosis across datasets in this study

From: Enhancing diagnostic deep learning via self-supervised pretraining on large-scale, unlabeled non-medical images

Labels

VinDr-CXR

ChestX-ray14

CheXpert

MIMIC-CXR

UKA-CXR

PadChest

Cardiomegaly

✔

✔

✔

✔

✔

✔

Pleural effusion

✔

✔

✔

✔

 

✔

Pleural effusion right

    

✔

 

Pleural effusion left

    

✔

 

Pleural thickening

✔

✔

   

✔

Infiltrates

     

✔

Pneumonia

✔

✔

✔

✔

 

✔

Pneumonia right

    

✔

 

Pneumonia left

    

✔

 

Pneumothorax

✔

✔

✔

✔

 

✔

Atelectasis

✔

✔

✔

✔

 

✔

Atelectasis right

    

✔

 

Atelectasis left

    

✔

 

Consolidation

✔

✔

✔

✔

 

✔

Congestion

    

✔

✔

Nodule/mass

✔

    

✔

Nodule

 

✔

    

Mass

 

✔

    

Fibrosis

✔

✔

    

Hernia

 

✔

   

✔

Emphysema

 

✔

   

✔

Edema

 

✔

    

Aortic elongation

     

✔

Kyphosis

     

✔

COPD

     

✔

Scoliosis

     

✔

Lung opacity

✔

 

✔

✔

  

Lung lesion

  

✔

✔

  

Fracture

  

✔

✔

  

No finding (healthy)

✔

✔

✔

✔

✔

✔

  1. The table details the specific labels applied to each dataset’s images for diagnostic purposes. The study’s multilabel diagnosis tasks involved predicting 11, 14, 10, 10, 9, and 17 distinct labels for the VinDr-CXR, ChestX-ray14, CheXpert, MIMIC-CXR, UKA-CXR, and PadChest datasets, respectively. Notably, UKA-CXR delineates separate labels for the presence of atelectasis, pleural effusion, and pneumonia for both the right and left sides of the chest. The “Healthy” label signifies cases without any disease diagnosis. ✔ label utilized in this study
  2. COPD Chronic obstructive pulmonary disease
Back to article page