RESUMO
OBJECTIVE. The purpose of our study was to assess whether linear ventricular dimensions-specifically, the frontal occipital horn ratio (FOHR) and frontal temporal horn ratio (FTHR) obtained from ultrasound (US)-are reliable measures of ventriculomegaly in infants. Our hypothesis was that these US measures would show good correlation with linear ventricular indexes and ventricular volumes obtained from MRI. MATERIALS AND METHODS. We retrospectively identified 90 infants (age ≤ 6 months corrected gestational age) with ventriculomegaly from 2014 to 2017 who had a total of 100 sets of US and MRI studies performed in a 3-day period. FOHR and FTHR were independently measured on US and MRI by two pediatric radiologists and two pediatric neuroradiologists, respectively. Ventricular and brain volumes were segmented from the MR images, and the ventricle-to-intracranial volume ratio was calculated. MRI served as the reference standard. Intraclass correlation coefficients and Bland-Altman analyses were generated to evaluate interobserver and US-MRI concordance. We assessed correlation of the FOHR and FTHR with the ventricle-to-intracranial volume ratio. RESULTS. Bland-Altman plots of the FOHR and FTHR between US and MRI showed excellent concordance with a bias of 0.05 (95% CI, -0.04 to 0.14) and 0.03 (95% CI, -0.06 to 0.13), respectively. There was good-to-excellent interobserver concordance for FOHR and FTHR on head US or MRI (r = 0.86-0.96). There was good correlation between ventricle-to-intracranial volume ratios and US- and MRI-derived FOHRs and FTHRs (r = 0.79-0.87). CONCLUSION. FOHR and FTHR obtained from US in infants with ventriculomegaly have excellent interobserver concordance, are concordant with MRI-derived linear ratios, and correlate with MRI-derived ventricular volumes. Therefore, US-derived FOHR and FTHR are reliable indexes for clinical follow-up of infantile ventriculomegaly.