Attend-and-Refine: Interactive keypoint estimation and quantitative cervical vertebrae analysis for bone age assessment

Open Webpage

Jinhee Kim, Taesung Kim, Taewoo Kim, Dong-Wook Kim, Byungduk Ahn, Yoon-Ji Kim, In-Seok Song, Jaegul Choo

Published: 2025, Last Modified: 15 Nov 2025Medical Image Anal. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•ARNet minimizes keypoint annotation effort with interactive estimation.•Proposed approaches incorporate a gating and cross–attention mechanism to better propagate user feedback.•Morphology-aware loss ensures structural consistency of keypoints.•Growth potential is estimated via quantitative cervical vertebrae morphology analysis.•AI tool aids clinicians in determining the optimal timing for orthodontic treatments.