Kim Annink

259 General discussion / Conclusions & clinical implications / Future directions neurodevelopmental outcome in a large cohort. To be able to score the MB, DWI and T1- and T2-weighted MRI should have a maximum slice thickness of 2mm. When using thicker slices, the MB are often missed. Additionally, the scoring system of Weeke et al. also includes an abnormal signal of the cerebellum on diffusion weighted MRI. Based on chapter two, one should realize when using this scoring system that even when the cerebellum appears normal on diffusion weighted MRI, there still might be (histopathological) cerebellar injury. It would be interesting to score the vermis separately from the cerebellar hemispheres, since ADC values in the vermis seem most reliable. Other MRI sequences, as magnetic resonance spectroscopy (MRS) and DTI might be more accurate in diagnosing cerebellar injury (27). The hippocampus, cerebellum and MB were not included in the CUS scoring system in chapter four , because the quality of CUS was suboptimal to score these structures and the cerebellar view was often not present. The quality and possibilities of CUS increases rapidly by the use of new techniques, potentially allowing inclusion of these structures in the CUS scoring system in the future as well. 12