Local medical data were collected from the University Hospital in Batna. After filtering and checking each image, we obtained 85 images capable of being examined by a doctor to help diagnose diseases.
As a result, After discussion with the medical specialists, we have decided to alter our approach to focus on identifying disease syndromes in X-ray images, rather than pinpointing the specific disease itself.
The diseases were diagnosed by Doctor M.Ahmed Azi's assistant professor at the pneumo- physiologic department, and Doctor Gadda Atef an assistant professor in pediatric anesthesia and resuscitation
The data contained 13 classes : 'Atelectasis', 'Cardiomegaly', 'Effusion', 'Infiltration', 'Nodule', 'Pneumonia', 'Pneumothorax', 'Emphysema', 'Fibrosis', 'Hernia'