Purpose or Case Report: Neck/airway radiographs in children are difficult due to variable soft tissues from swallowing/crying and positioning; short necks and limited cooperation further impact quality. No pediatric image-quality or technique guidelines exist. We aimed to develop a quantitative image-quality assessment tool.
Methods & Materials: IRB-exempt quality-improvement project. Two pediatric radiology attendings, one fellow, one physicist, and two technologists reviewed 30 exams (lateral neck [L], anteroposterior airway [AP]) from August 2025 and defined failure modes. Interrater agreement was measured in a teaching session (six residents, four fellows): 15-minute instruction, three consensus examples (adequate/inadequate), then five 2-view cases rated independently. Agreement used Gwet’s AC1. Training-level effects used the Fisher–Tukey method (p<0.05). To derive a composite score, failure modes were rank-ordered and weighted by clinical relevance and likelihood of repeat imaging if inadequate.
Results: Failure modes were similar for L and AP. AC1: anatomic landmark visibility (range 0.42–0.77 [L], 0.77–0.85 [AP]), patient positioning (0.35 [L], 0.21 [AP]), beam penetration (0.59 [L], 0.68 [AP]), external artifacts (0.77 [L], 0.86 [AP]), equipment artifacts (0.84 [L], 0.83 [AP]), motion (0.77 [L], 0.85 [AP]), field of view (0.33 [L], 0.47 [AP]), labeling (0.89 [L], 1.00 [AP]). Fellows more often rated images as adequate for palatine tonsils (p=0.027), epiglottis (p=0.033), and subglottic airway (p=0.013); no other differences were significant. Highest agreement was AP labeling (AC1=1.00); lowest was AP positioning (AC1=0.21). Rank order/weights: Lateral—anatomic landmarks 0.45 (retropharynx 0.17; epiglottis 0.11; tonsils 0.08), positioning 0.25 (chin straight/30° 0.18; mouth closed/no pacifier 0.07), external artifacts 0.10, beam penetration 0.07, motion 0.04, equipment artifacts 0.04, field of view 0.04, labeling 0.01. AP—anatomic landmarks 0.40 (subglottic airway 0.35; glottis 0.04; bronchial bifurcation 0.01), positioning 0.30 (neck extension), field of view 0.15 (inferior 0.11; R–L 0.02; superior 0.02), beam penetration 0.10, external artifacts 0.02, motion 0.01, equipment artifacts 0.01, labeling 0.01.
Conclusions: A standardized image-quality tool for pediatric neck/airway radiographs is feasible and reproducible. Moderate interrater agreement supports reliability, with greatest variability in positioning and field of view. Wider use may improve consistency and diagnostic quality.