Application of three-dimensional visualization reconstruction technology through CT to quantitatively measure diastasis recti abdominis

Background Diastasis recti abdominis (DRA) is characterized by widening of the linea alba without a true fascial defect, but conventional ultrasound primarily provides linear measurements and two-dimensional CT approaches commonly rely on geometric assumptions for volumetric estimates. Whether CT-based three-dimensional visualization reconstruction (3DVR) provides additional quantitative characterization and systematically different measurements remains uncertain. We therefore compared quantitative measurements of DRA obtained using CT-based 3DVR, two-dimensional CT ellipsoid approximation (2D-EA), and ultrasound (US) in an exploratory methodological study. Methods Twenty patients with DRA were included in this exploratory observational methodological comparison. CT-based 3DVR was used to quantify inter-rectus distance (IRD), defect surface area (DSA), non-hernial abdominal wall bulging-space volume (HSV; abbreviation retained from the original dataset), abdominal cavity volume (ACV), and loss of domain (LOD). A two-dimensional CT ellipsoid approximation (2D-EA) was used for comparison, and all participants also underwent ultrasound measurement of IRD. LOD was calculated as HSV/(HSV + ACV) × 100%. Normality of paired differences was evaluated using the Shapiro-Wilk test. Paired t-tests were used when normality was supported, with Wilcoxon signed-rank tests as sensitivity analyses. Exact P values, 95% confidence intervals, Cohen’s dz effect sizes, Pearson correlations, and a Bonferroni-adjusted significance threshold of P < 0.0071 were reported. Results 3DVR yielded higher ACV than 2D-EA (5340.51 ± 818.53 vs. 4497.82 ± 932.28 cm³; mean paired difference, 842.69 cm³; 95% CI, 577.01-1108.36; P < 0.001; Cohen’s dz = 1.484), higher HSV (248.71 ± 128.95 vs. 190.28 ± 107.05 cm³; difference, 58.43 cm³; 95% CI, 27.53–89.33; P = 0.0008; dz = 0.885), and higher IRD (7.88 ± 1.50 vs. 7.59 ± 1.40 cm; difference, 0.29 cm; 95% CI, 0.11–0.47; P = 0.0035; dz = 0.745). LOD did not differ significantly (4.37 ± 2.05% vs. 3.90 ± 1.69%; P = 0.0886). Mean DSA measured by 3DVR was 115.15 ± 27.79 cm². IRD was also higher with 3DVR than with ultrasound (7.88 ± 1.50 vs. 6.82 ± 1.16 cm; difference, 1.06 cm; 95% CI, 0.40–1.72; P = 0.0034; dz = 0.747), whereas the 2D-EA-versus-ultrasound difference did not meet the Bonferroni-adjusted threshold ( P = 0.0314). The paired differences were compatible with normality, and the principal findings were unchanged using the Bonferroni-adjusted threshold of P < 0.0071. Conclusion CT-based 3DVR demonstrated systematic differences in volumetric and linear measurements compared with 2D-EA and ultrasound and provided DSA, which was not obtainable using the ellipsoid approximation. These findings suggest that 3DVR may offer additional three-dimensional anatomical characterization when CT images are available from appropriately justified examinations; however, the approximately 6-hour reconstruction time per case, lack of a reference standard, and absence of clinical-outcome validation currently limit routine implementation. Future studies should prioritize workflow automation, standardized multimodal acquisition, reproducibility testing, and prospective evaluation of whether 3D-derived measurements influence treatment planning or outcomes. Clinical trial number Not applicable.