[P2-169] T2 Quantification MRI Predicts Cerebral Palsy in Preterm Infants
Introduction: Our study aims to predict neurological outcomes with T2 quantification MRI among preterm infants.
Methodology: During January 1st 2012 to May 31st 2015, preterm infants with T2 quantification brain MRI before 1 year old were retrospectively reviewed. Patients with neurodevelopmental follow-up for more than 12 months were included. Those with major neurological underlying diseases were excluded. Values of T2 quantification were measured over 3 sections of periventricular and 4 lobes of subcortical white matters. T2 quantification in relation to corrected age were analyzed with restricted cubic spline regression. Prediction of CP was examined with receiver operating characteristic curve, and assessed with improvement discrimination index and net reclassification index. The institutional review board of National Cheng Kung University Hospital approved the study.
Results: During the study period, 57 preterm infants underwent T2 quantification MRI. After excluding 19 patients, 38 preterm infants were enrolled for analysis. Twenty patients (52.6%) had neurodevelopmental abnormality, including 8 (21%) of DD without CP, and 12 (31.6%) of CP. The T2 quantification in relation to age was curvilinear in normal-development preterm infants, linear in CP without DD, and flat in CP. When the MRI was performed at > 1 month old of corrected age, CP could be predicted with T2 quantification measured over sections through frontal horns and mid-body periventricular white matter (Area under the ROC curve 0.936).
Conclusions: T2 quantification MRI in premature infants provides prognostic prediction on neurodevelopmental outcomes. Age-dependent and area-selective interpretation in preterm brains should be emphasized.
Methodology: During January 1st 2012 to May 31st 2015, preterm infants with T2 quantification brain MRI before 1 year old were retrospectively reviewed. Patients with neurodevelopmental follow-up for more than 12 months were included. Those with major neurological underlying diseases were excluded. Values of T2 quantification were measured over 3 sections of periventricular and 4 lobes of subcortical white matters. T2 quantification in relation to corrected age were analyzed with restricted cubic spline regression. Prediction of CP was examined with receiver operating characteristic curve, and assessed with improvement discrimination index and net reclassification index. The institutional review board of National Cheng Kung University Hospital approved the study.
Results: During the study period, 57 preterm infants underwent T2 quantification MRI. After excluding 19 patients, 38 preterm infants were enrolled for analysis. Twenty patients (52.6%) had neurodevelopmental abnormality, including 8 (21%) of DD without CP, and 12 (31.6%) of CP. The T2 quantification in relation to age was curvilinear in normal-development preterm infants, linear in CP without DD, and flat in CP. When the MRI was performed at > 1 month old of corrected age, CP could be predicted with T2 quantification measured over sections through frontal horns and mid-body periventricular white matter (Area under the ROC curve 0.936).
Conclusions: T2 quantification MRI in premature infants provides prognostic prediction on neurodevelopmental outcomes. Age-dependent and area-selective interpretation in preterm brains should be emphasized.