Validating conventional 2D IMPA against 3D one: is it still a trustworthy angle? A retrospective CBCT-based diagnostic agreement study

Aim: This study tested whether conventional 2D incisor mandibular plane angle (IMPA) can be used interchangeably with a rigorously defined 3D one. Material and methods: This single-centre retrospective study included 169 CBCT scans. Synthetic lateral cephalograms were generated from DICOM data to calculate the conventional 2D IMPA between the Go–Me mandibular plane and the lower central incisor long axis. The 3D IMPA used the axes of teeth 3.1 and 4.1 projected three-dimensionally to the orthogonal plane to GoL-GoR-Me and the higher angle was selected. Normality of paired differences was assessed using the Shapiro-Wilk test. Systematic bias was evaluated with a paired t-test. Association between 2D and 3D IMPA was assessed using Pearson's correlation and linear regression. Agreement was quantified by Bland-Altman analysis. Classification agreement within the 85–95° range was evaluated using Cohen's kappa and McNemar's test. Results: Mean 2D and 3D IMPA were 95.29 ± 8.98° and 95.75 ± 9.42°, respectively, with a mean paired 3D-2D difference of 0.46° (paired t-test, P = 0.084). Correlation was strong (r = 0.931), and linear regression yielded R2 = 0.867. Bland-Altman bias was 0.46° (95%LoA −6.29° to +7.21°). Classification agreement across the 85–95° range was 83.43% (κ = 0.636; McNemar P = 0.345). Conclusions: 2D and 3D IMPA showed negligible mean bias, excellent association, and substantial diagnostic concordance. Routine cases may rely on 2D IMPA without material loss of information, whereas in torque-critical, anatomically constrained, or borderline cases the observed ±7° dispersion argues for 3D assessment.