Thin-slice MRI versus transvaginal ultrasound and sonohysterography for uterine niche evaluation: a reproducibility study

Ultrasound and magnetic resonance imaging (MRI) are widely used to evaluate uterine niches, yet the reproducibility of their quantitative measurements remains uncertain. This study aimed to compare niche metrics obtained via transvaginal ultrasound (TVS), sonohysterography (SHG), and thin-slice MRI in women with hysteroscopically confirmed niches, assessing measurement reliability across modalities. This prospective observational study enrolled 60 symptomatic women with hysteroscopically confirmed uterine niches between June 2024 and February 2025. Preoperatively, all participants underwent TVS, SHG, and thin-slice MRI (1-mm slice thickness). Independent observer pairs reviewed the ultrasound and MRI datasets for niche detection and quantified niche length, depth, residual myometrial thickness (RMT), and total area. Intra-observer, inter-observer, and inter-method agreements were evaluated using intraclass correlation coefficients (ICCs) and Bland-Altman analysis. RMT measurements varied significantly by modality: 3.15 ± 1.57 mm for TVS, 2.69 ± 1.33 mm for SHG, and 2.39 ± 1.24 mm for thin-slice MRI ( P < 0.01). Thin-slice MRI yielded the highest intra-observer reliability (ICC: 0.93–0.99), followed by SHG (ICC: 0.80–0.87) and TVS (ICC: 0.64–0.75). Inter-observer reliability followed a similar pattern, remaining excellent for thin-slice MRI (ICC: 0.89–0.99) but moderate for SHG (ICC: 0.43–0.60). Agreement between SHG and thin-slice MRI proved poor (ICC < 0.40) across all parameters. Furthermore, MRI outperformed ultrasound in identifying niche-associated microstructures, capturing more lateral branches (8/10) and cysts (11/13). Thin-slice MRI demonstrated superior measurement consistency relative to TVS and SHG within this symptomatic, surgically selected cohort. While ultrasound remains the primary modality, thin-slice MRI offers essential anatomical precision for complex niches or pre-surgical planning when initial ultrasound findings are equivocal.