See Hamamci et al., MICCAI 2023
Numerous Machine Learning models for the interpretation of panoramic dental X-rays have been developed, yet none of them offers an end-to-end solution that identifies problematic teeth with dental enumeration and associated diagnoses at the same time. In this work, the Menze group structure three distinct types of annotated data hierarchically following the FDI system. To learn from all three hierarchies jointly, a novel diffusion-based hierarchical multi-label object detection framework is introduced by adapting a diffusion-based method that formulates object detection as a denoising diffusion process from noisy boxes to object boxes. Experimental results show that this novel method significantly outperforms state-of-the-art object detection methods.
The code and the datasets are available at https://github.com/ibrahimethemhamamci/HierarchicalDet.