A novel deep learning architecture, the “relational reasoning network,” can learn the spatial relationships between key anatomical landmarks of craniomaxillofacial bones from computed tomography images. (IMAGE)
Caption
A novel deep learning architecture, the “relational reasoning network,” can learn the spatial relationships between key anatomical landmarks of craniomaxillofacial bones from computed tomography images. This enables a trained model to find these landmarks automatically and accurately in images from all types of patients, even those with deformities that would trip up other deep learning models.
Credit
Credit: Torosdagli et al., doi 10.1117/1.JMI.10.2.024002.
Usage Restrictions
NONE
License
Original content