Brainstem–vermis and brainstem–tentorium angles: 3D ultrasound study of the intra- and inter-observer agreement

Abstract

Objective: The brainstem–vermis angle (BVA) and the brainstem–tentorium angle (BTA) have been proposed to quantify vermian and tentorial alterations associated to fetal posterior fossa malformations. Our objective is to evaluate the intra- and inter-observer agreement for the measurement of these angles during the prenatal ultrasound scan. Methods: Fifteen 3 D sets from fetuses with posterior fossa pathology were processed by multiplanar navigation. Four experts in prenatal ultrasound measured the BVA and the BTA on a mid-sagittal view of the fetal brain, following the criteria described by Volpe. Observer 1 did also repeated measurements following both the criteria by Volpe and Ghi. The intraclass correlation coefficient (ICC) was used to test consistency and absolute agreement for intra- and inter-observer measurements. The concordance is poor when ICC <0.40 and excellent when ICC >0.75. Results: Consistency ICC for measurements between the 4 observers was 0.74 (95% CI: 0.54–0.89) for the BVA and 0.55 (95% CI: 0.29–0.78) for the BTA. Absolute agreement ICC was 0.67 (BVA) and 0.45 (BTA). The intra-observer study showed a significantly greater angle with the criteria by Ghi (mean 52.1°, 95% CI: 49.5–54.8) than with the criteria by Volpe (mean 37.4°, 95% CI: 33.1–41.6). Conclusions: A more precise methodology for the measurement of these angles should be described.