Abstract:Clinicians trace cephalometric radiographs by following a structured anatomical workflow -- yet no prior system explicitly encodes this into computation. We present a five-phase anatomy-guided pipeline producing confidence-weighted spatial priors that shape HRNet-W32 training. The system achieves 1.04 mm mean radial error on 25 landmarks across 1,502 radiographs from 7+ imaging devices -- comparable to HYATT-Net (1.05 mm on CEPHA29) via explicit anatomical priors rather than learned attention. A three-way ablation isolates the mechanism: anatomical priors maintain a 1% validation-to-test gap, while removing priors yields an 88% gap (1.94 mm) -- despite identical validation convergence. A training x inference prior matrix confirms that (1) all models are inference-independent, (2) the 28-channel architecture alone provides no benefit, (3) random priors are partial and unstable (1.72 mm), and (4) only anatomically correct, image-specific priors yield 1.04 mm -- functioning as a training-time regularizer. No prior generation is needed at deployment. Five-fold cross-validation (p=0.0015), patient-level permutation testing (p<0.0001, n=151), reproduced baselines, Grad-CAM analysis, and clinical validation (100% skeletal classification across 151 patients including 72 boundary cases, kappa=1.00) provide converging evidence. Cross-domain experiments support the hypothesis that prior effectiveness depends on landmark spatial entropy -- confirmed prospectively across four domains. Supplementary materials included.
| Comments: | v2: Added patient-level permutation test (p<0.0001), hand X-ray prospective SEI validation with held-out test results, CSXA test-set evaluation, training x inference prior matrix, claim discipline table. Supplementary materials included. 20 pages, 23 tables, 15 figures, 35 references |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2605.03358 [cs.CV] |
| (or arXiv:2605.03358v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2605.03358 arXiv-issued DOI via DataCite |
From: Sidhartha Mohapatra [view email]
[v1]
Tue, 5 May 2026 04:33:45 UTC (4,731 KB)
[v2]
Sun, 24 May 2026 05:06:21 UTC (13,556 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。