Multi-label Classification of Pathologies in Chest Radiograph Images Using DenseNet

• Alison Corrêa Mendes UFMA https://orcid.org/0009-0004-3687-5519
• Alexandre César Pinto Pessoa UFMA https://orcid.org/0000-0003-4995-8909
• Anselmo Cardoso de Paiva UFMA https://orcid.org/0000-0003-4921-0626

Resumo

Chest radiography exams are still one of the main methods for detecting and diagnosing certain thoracic pathologies. This study evaluates the performance of a DenseNet in a multi-label classification task on radiography images, using focal loss as the loss function to address the class imbalance problem. For the experiments, 14 different types of findings were considered. Satisfactory results were obtained using the area under the ROC curve (AUC-ROC) as the metric, where the average performance across all classes was 0.861.