Predicting when dental growth spurts occur has long challenged clinicians, as treatment before or after a growth peak can be less effective. Now, researchers in South Korea have developed an artificial intelligence (AI) system that can forecast those peaks using a simple neck X-ray.
Researchers from Korea University Anam Hospital, KAIST and the University of Ulsan created an AI model called Attend-and-Refine Network (ARNet-v2) to identify puberty-related growth changes from a single lateral cephalometric radiograph. The study, led by Dr. Jinhee Kim and Prof. In-Seok Song, was published July 29, 2025, in Medical Image Analysis (Vol. 106, December 2025).
Related: AI transcription tools could streamline dental record-keeping, study finds—but caution urged
Trained on more than 5,700 radiographs and validated across four public medical-imaging datasets, ARNet-v2 outperformed previous systems, reducing prediction failures by up to 67 per cent and cutting the number of manual corrections in half. Its interactive design allows a clinician’s single adjustment to automatically refine related anatomical points, improving both speed and accuracy.
“Clinically, the model’s ability to extract precise cervical-vertebra keypoints from one X-ray enables accurate estimation of a child’s pubertal growth peak, a key factor in determining the timing of orthodontic treatment,” Prof. Song said. “By replacing traditional hand-wrist radiographs, it can lower radiation exposure and costs for young patients.”
Related: The consent gap: What the Heartland lawsuit teaches us about AI in dentistry
Because the algorithm relies on a single radiograph, it reduces the need for additional imaging and lowers the cost of manual annotation. The same AI framework may also be applied to other medical-imaging fields such as brain MRI, retinal scans and cardiac ultrasound, and even non-medical areas like robotics and autonomous driving.
Researchers say ARNet-v2 could make growth assessment more efficient in hospitals and remote clinics alike, potentially making AI-assisted bone-age analysis a standard component of paediatric orthodontics.
Source link