(1111) Amniotic Fluid Volume Estimation: A Novel Approach for an Old Problem

46 singleton pregnancies (median GA 38.4 weeks) were analyzed. Mean AFI was 12.2±4.5 cm, SDP 5.1±2.0 cm, true AFV 254±176 mL, and 3D4Q 220±155 mL. Mean Q (54±56 mL; 0–287 mL) correlated best with reference AFV (ρ=0.88, p< 10⁻¹⁴), followed by 3D4Q (ρ=0.87) and Qmax (ρ=0.76); conventional 2D indices had weaker correlation (SDP ρ=0.62, AFI ρ=0.55). Mean Q showed the highest correlation (ρ = 0.88, p < 1×10⁻¹⁴), closely followed by the 3D4Q (ρ = 0.87, p < 1 × 10⁻¹⁴). The Q max was also strongly related to true AFV (ρ=0.76, p < 1 × 10⁻⁹). By contrast, conventional 2D indices had moderate associations- SDP: ρ = 0.62 (p ≈ 4×10⁻⁶), AFI: ρ = 0.55 (p ≈ 6×10⁻⁵). Multiple ordinary-least-squares regressions for candidate predictors are shown in Table 1, and the best fit was the formula comprising of measurements of all 4 quadrants and Qmax. Bland–Altman analysis showed minimal bias (~–1 mL) for all models, with the KUGOAF-Qmax yielding ±140 mL limits of agreement, the two-parameter KUGOAF-MeanQmax slightly narrower (±133 mL), and Qmax alone wider (±182 mL), indicating increased random error in the single-predictor model.


Conclusion:

The mean 3D volume of all quadrants had the strongest correlation with the real AFV, while 2D SDP and AFI had weaker correlations. Multi-pocket formulas demonstrated clinically acceptable agreement within ±140 mL.