Inter-site generalizability of EEG based age prediction algorithms in the preterm infant.

Objective. To overcome the effects of site differences in EEG-based brain age prediction in preterm infants.Approach. We used a 'bag of features' with a combination function estimated using support vector regression (SVR) and feature selection (filter then wrapper) to predict post-menstrual age (PMA). The SVR was trained on a dataset containing 138 EEG recordings from 37 preterm infants (site 1). A separate set of 36 EEG recordings from 36 preterm infants was used to validate the age predictor (site 2). The feature distributions were compared between sites and a restricted feature set was constructed using only features that were not significantly different between sites. The mean absolute error between predicted age and PMA was used to define the accuracy of prediction and successful validation was defined as no significant differences in error between site 1 (cross-validation) and site 2.Main results. The age predictor based on all features and trained on site 1 was not validated on site 2 (p< 0.001; MAE site 1 = 1.0 weeks,n= 59 versus MAE site 2 = 2.1 weeks,n= 36). The MAE was improved by training on a restricted features set (MAE site 1 = 1.0 weeks,n= 59 versus MAE site 2 = 1.1 weeks,n= 36), resulting in a validated age predictor when applied to site 2 (p= 0.68). The features selected from the restricted feature set when training on site 1 closely aligned with features selected when trained on a combination of data from site 1 and site 2.Significance. The ability of EEG classifiers, such as brain age prediction, to maintain accuracy on data collected at other sites may be challenged by unexpected, site-dependent differences in EEG signals. Permitting a small amount of data leakage between sites improves generalization, leading towards universal methods of EEG interpretation in preterm infants.

Group B Streptococcus colonization at delivery is associated with maternal peripartum infection.

BACKGROUND: Group B Streptococcus (GBS) is a major cause of serious neonatal infection but its role in maternal morbidity has received little investigation. The aim of this study was to determine whether GBS colonization at delivery is associated with increased risk of maternal peripartum infection. METHODS: In this prospective cohort study, 1746 unselected women had a vaginal-rectal culture taken at the onset of labor. Diagnosis of maternal peripartum infection was based on a combination of two or more signs or symptoms including fever, breast pain, severe wound or pelvic pain, purulent discharge and abnormal laboratory tests including C-reactive protein and white blood cell count occurring from labor until 2 weeks postpartum. The main outcome measure was the proportion of women with maternal peripartum infection according to GBS colonization status. RESULTS: A total of 25.9% (452/1746) women were colonized with GBS. The rate of peripartum infection was almost twice as high in colonized women (49/452 [10.8%]) vs. non-colonized women (81/1294 [6.3%]); OR 1.82 [1.26-2.64], p = 0.002). This association was confirmed in a multivariable model (OR 1.99 [1.35-2.95], p = 0.001). Women diagnosed with peripartum infection had a significantly longer hospital stay compared to women without peripartum infection (4 days (median) vs. 3 days, p < 0.001). Length of hospital stay did not differ between colonized and non-colonized women. Serotype IV GBS was more frequent in colonized women with peripartum infection than in women without peripartum infection (29.3% vs. 12.5%, p = 0.003). CONCLUSIONS: GBS colonization at delivery is associated with increased risk of peripartum infection. Whether this increase is due directly to invasion by GBS or whether GBS colonization is associated with a more general vulnerability to infection remains to be determined.

Preserved endothelial function in young adults with type 1 diabetes.

BACKGROUND AND AIM: Endothelial dysfunction is involved in the pathogenesis of atherosclerosis and is typically present in older adults with type 1 diabetes (T1D). In young adults, we aimed to assess the impact of T1D on endothelial function as detected by digital peripheral arterial tonometry (PAT) and its relationship with cardiovascular risk factors and long term glycemic control. MATERIALS AND METHODS: Reactive hyperemia index (RHI) as a measure of endothelial function was assessed by PAT in 46 T1D patients and 32 healthy controls. All were participants in the "Atherosclerosis and Childhood Diabetes" study, with baseline values registered five years previously. Annual measurements of HbA1c for assessment of glycemic burden were provided by the Norwegian Childhood Diabetes Registry. RESULTS: The diabetes patients had a mean age of 20.8 years, a median duration of diabetes of 10.0 years and a mean HbA1c of 8.7%. RHI was not significantly decreased in the diabetes group, mean 2.00 (SD = 0.59) vs. 2.21 (SD = 0.56), p = .116. There was no gender difference or any associations with traditional risk factors. Furthermore, there was no significant association between RHI and either HbA1c or long term glycemic burden. CONCLUSIONS: RHI as a measure of endothelial function was preserved in young adults with T1D compared with healthy controls.