Right and left ventricular 24‐segment sphericity index is abnormal in small‐for‐gestational‐age fetuses

Objective Fetuses with growth restriction have been reported to have an abnormal sphericity index (SI), which is indicative of the shape of the ventricular chambers of the heart. Our aim was to evaluate the SI for 24 transverse segments distributed from base to apex of the right (RV) and left (LV) ventricles to determine whether, in small‐for‐gestational‐age (SGA) fetuses, the SI is abnormal at locations other than the basal segment. Methods We evaluated 30 SGA fetuses between 25 and 37 weeks of gestation. SI was computed for both ventricles by dividing the end‐diastolic mid‐basal–apical length by each of 24 end‐diastolic transverse segmental widths, from base (Segment 1) to apex (Segment 24). For each ventricle, the Z ‐score and centile for the SI from each of the 24 segments were computed using the mean and SD from published equations. The 24‐segment method, defining abnormal SI as values < 10 th centile or > 90 th centile, was compared with that of using only the basal segment by chi‐square analysis to determine the number of fetuses identified with an abnormal SI. Results In 23 of the 30 (77%) SGA fetuses, at least one of the 24 transverse segments in one or both ventricles had an abnormal SI; in 17% of cases, both ventricles were affected, in 23% of cases only the RV was involved and in 37% of cases only the LV was involved. Compared with the 24‐segment model, significantly fewer fetuses with an abnormal SI were identified using only basal Segment 1, from the RV base (58%, 7/12; P < 0.01) or only Segment 12, in the mid portion of the RV (50%, 6/12; P < 0.005). Combining measurements of Segment 1 and Segment 12 from the RV identified 83% of fetuses with at least one abnormal SI and was not significantly different from using the 24‐segment model. Similarly, significantly fewer fetuses with an abnormal SI were identified using only LV basal Segment 1 (63%, 10/16; P < 0.006) or only Segment 12, in the mid portion of the LV (75%, 12/16; P < 0.03), when compared with the 24‐segment model. Combining measurements of both LV Segment 1 and Segment 12 identified 81% (13/16) of fetuses with an abnormal SI and was not significantly different from using the 24‐segment model. Conclusion The 24‐segment SI of RV and LV provides a comprehensive method with which to examine the shape of the ventricular chambers and identifies more SGA fetuses with an abnormal SI than are identified using only the basal segment SI.