Computer-Aided Diagnosis for Pneumoconiosis Staging Based on Multi-scale Feature Mapping

Abstract In this research, we explored a method of multi-scale feature mapping to pre-screen radiographs quickly and accurately in the aided diagnosis pneumoconiosis staging. We utilized an open dataset self-collected as research datasets. proposed model based on deep learning extraction technology for detecting pulmonary fibrosis discrimination The diagnostic accuracy was evaluated using under curve (AUC) receiver operating characteristic (ROC) curve. AUC value our 0.84, which showed best performance compared with previous work results indicated that highly consistent clinical experts real patient. Furthermore, obtained through categories I–IV testing demonstrated I (AUC = 0.86) IV 0.82) achieved level clinician categorization. Our could be applied pre-screening clinic.