AdURA-Net: Adaptive Uncertainty and Region-Aware Network

AdURA-Net is a new AI framework designed for medical image analysis that addresses uncertainty in clinical decision-making for thoracic disease classification. The system uses adaptive dilated convolution and a dual head loss function to handle uncertain diagnostic labels in medical datasets like CheXpert and MIMIC-CXR.

• →AdURA-Net addresses the critical problem of uncertainty in automated medical diagnosis where confident predictions may not be appropriate.
• →The framework combines adaptive dilated convolution with multiscale deformable alignment to capture anatomical complexities in medical images.
• →A dual head loss function integrates masked binary cross entropy with Dirichlet evidential learning for better uncertainty handling.
• →The system is specifically designed for multilabel medical datasets that include positive, negative, and uncertain classifications.
• →The research focuses on enabling AI models to acknowledge when they lack sufficient evidence for confident medical predictions.