Egyptian Journal of Radiology and Nuclear Medicine · Published 2026-08-05 · DOI 10.1186/s43055-026-01831-y
Islam Haney Shawali, Karthick Raja Kasirajan, Mohamed Elsayed Elshaaer, Maryse Awadallah
Abstract Background Accurate preoperative quantification of loss of domain (LOD) is essential in patients with incisional hernia because it influences surgical planning and predicts the need for adjunctive preoperative optimization. Although three-dimensional (3D) CT volumetric reconstruction is considered anatomically more precise, two-dimensional (2D) CT geometric estimation is simpler, faster, and universally available. This study compared both techniques to determine whether 2D assessment provides clinically acceptable agreement with 3D volumetry. Objectives In this retrospective study, 50 adults with CT-confirmed incisional hernia underwent LOD assessment using two approaches: (1) ellipsoid-based 2D geometric estimation from linear CT measurements and (2) semi-automated 3D volumetric segmentation. Mean abdominal cavity volume, hernia sac volume, and LOD were compared using paired t-tests, Pearson correlation, Bland–Altman agreement analysis, and intraclass correlation coefficients (ICC). Methods Mean abdominal cavity volume was 6690 ± 2054 cm 3 using the 2D method and 6898.9 ± 2122.8 cm 3 using the 3D method (p = 0.34). Mean hernia sac volume was 1375 ± 953 cm 3 and 1381 ± 947 cm 3 , respectively (p = 0.81). Corresponding mean LOD values were 20.6 ± 12.38% and 16.8 ± 8.72%, with no statistically significant difference (p = 0.13). The two techniques demonstrated a strong positive correlation (r = 0.83, p < 0.001). Bland–Altman analysis showed a mean bias of 3.8% with limits of agreement ranging from − 4.2 to + 11.8%, indicating acceptable agreement at the population level but clinically relevant variability in individual patients. Interobserver reproducibility was excellent (overall ICC = 0.91). Results Mean 2D LOD was 20.6 ± 12.38%, while mean 3D LOD was 16.8 ± 8.72%, with no significant difference between techniques (p = 0.13). A strong correlation was observed between 2 and 3D LOD values (r = 0.83, p < 0.001). Bland–Altman analysis demonstrated a small positive bias of 3.8% with narrow limits of agreement (− 4.2– + 11.8%). Interobserver agreement was excellent for both methods (ICC = 0.91). Conclusion Two-dimensional CT geometric estimation demonstrated good agreement with 3D volumetric reconstruction and excellent reproducibility, supporting its use as a practical first-line technique for routine preoperative LOD assessment. However, the observed individual-level variability suggests that 3D volumetry remains preferable for complex or borderline cases in which small differences may influence surgical decision-making.
Abstract from DOAJ. Public domain (CC0 1.0).
Read the article at the publisher →
Shawali, I., Kasirajan, K., Elshaaer, M., et al. (2026). Comparative analysis of two-dimensional CT estimation versus three-dimensional volumetric reconstruction for loss-of-domain assessment in incisional hernia. Egyptian Journal of Radiology and Nuclear Medicine. https://doi.org/10.1186/s43055-026-01831-y