Kim Annink

53 Cerebellar injury in HIE: MRI versus histopathology normal postoperative MRI (range 3-13; median 7.5 days after birth) and normal neurodevelopmental follow-up at 18-24 months of age as a “control group”. Parental consent for postmortem histopathological examination of the brain was obtained for all patients. This study was approved by the medical ethical committee (#18-167) as a ‘non Medical Research Act’ study and the Biobank of the University Medical Center Utrecht approved to use the rest biomaterial for this research (biobank #18-284). Magnetic resonance imaging MRI was conducted in the first week after birth in the HIE group and within the first two weeks after birth in the control group using 1.5 or 3.0 Tesla (T) MRI scanners (Philips Medical Systems, Best, The Netherlands). Standard scan protocols included at least axial DWI and T1-weighted and T2-weighted images. ADC maps were created as described in Alderliesten et al. 2011 (10). The slice thickness was 4mm for both 1.5T and 3.0T MRI. ADC values were measured in the cerebellar hemispheres, the left and right DN and the vermis using Horos Imaging software (The Horos Project, available from: https://www.horosproject.org) . A region of interest (ROI) template was developed and copied to all scans (Figure 1), it was manually adapted per patient if necessary because of the anatomical variances between patients. The anatomy was verified using T1-weighted and T2-weighted MR images and inclusion of cerebrospinal fluid in the ROI was avoided at all times. ADC values were automatically measured within the ROI. Two independent raters measured ADC values in all patients to analyze interrater variability. 3