Automated Skeletal Classification with Lateral Cephalometry Based on Artificial Intelligence

• DOI: 10.1177/0022034520901715
• Publication Date: 2020-01-24T19:01:17Z
• Authors: H.J. Yu; S.R. Cho; M.J. Kim; W.H. Kim; J.W. Kim; J. Choi
• Keywords: Cephalogram; Cephalometry; Cephalometric analysis; Tracing

Lateral cephalometry has been widely used for skeletal classification in orthodontic diagnosis and treatment planning. However, this conventional system, requiring manual tracing of individual landmarks, contains possible errors inter- intravariability is highly time-consuming. This study aims to provide an accurate robust diagnostic system by incorporating a convolutional neural network (CNN) into 1-step, end-to-end with lateral cephalograms. A multimodal CNN model was constructed on the basis 5,890 cephalograms demographic data as input. The optimized transfer learning augmentation techniques. Diagnostic performance evaluated statistical analysis. proposed exhibited >90% sensitivity, specificity, accuracy vertical sagittal diagnosis. Clinical showed highest at 96.40 (95% CI, 93.06 98.39; III). receiver operating characteristic curve area under both demonstrated excellent mean >95%. heat maps were also provided deeper understanding quality learned visually representing region cephalogram that most informative distinguishing classes. In addition, we present broad applicability through subtasks. CNN-incorporated potential without need intermediary steps complicated procedures.