Cephalometric Landmark Detection on Cephalograms using Regression CNN

Cephalograms are an imaging process used by orthodontics to determine the position of cephalometric landmarks. These cephalometric landmarks are essentials to be found before any treatment to the patient can be performed. In this research, a CNN architecture to detect 31 cephalometric landmarks was proposed. The results shows that Mean Absolute Percentage Error (MAPE) of our proposed method achieves a value of 1.856%. Better than detectron2 with 2.443%.