Left atrial strain assessment unveils left ventricular diastolic dysfunction in neonates with transient tachypnea of the newborn: A prospective observational study.

INTRODUCTION: An inadequate clearance of lung fluid plays a key role in the pathogenesis of transient tachypnea of the newborn (TTN). OBJECTIVES: To evaluate if left ventricular diastolic dysfunction contributes to reduced clearance of lung fluid in TTN. MATERIALS AND METHODS: This was a prospective, observational study. Echocardiography and lung ultrasound were performed at 2, 24 and 48 h of life (HoL) to assess biventricular function and calculate lung ultrasound score (LUS). Left atrial strain reservoir (LASr) provided surrogate measurement of left ventricular diastolic function. RESULTS: Twenty-seven neonates with TTN were compared with 27 controls with no difference in gestation (36.1 ± 2 vs. 36.9 ± 2 weeks) or birthweight (2508 ± 667 vs. 2718 ± 590 g). Biventricular systolic function was normal in both groups. LASr was significantly lower in cases at 2 (21.0 ± 2.7 vs. 38.1 ± 4.4; p < 0.01), 24 (25.2 ± 4.5 vs. 40.6 ± 4.0; p < 0.01) and 48 HoL (36.5 ± 5.8 and 41.6 ± 5.2; p < 0.01), resulting in a significant group by time interaction (p < 0.001), after adjusting for LUS and gestational diabetes. A logistic regression model including LUS, birth weight and gestational diabetes as covariates, showed that LASr at 2 HoL was a predictor of respiratory support at 24 HoL, with an adjusted odds ratio of 0.60 (CI 0.36-0.99). CONCLUSIONS: LASr was reduced in neonates with TTN, suggesting diastolic dysfunction, that may contribute to the delay in lung fluid clearance.

Artificial Neural Networks for Short-Form Development of Psychometric Tests: A Study on Synthetic Populations Using Autoencoders.

Short-form development is an important topic in psychometric research, which requires researchers to face methodological choices at different steps. The statistical techniques traditionally used for shortening tests, which belong to the so-called exploratory model, make assumptions not always verified in psychological data. This article proposes a machine learning-based autonomous procedure for short-form development that combines explanatory and predictive techniques in an integrative approach. The study investigates the item-selection performance of two autoencoders: a particular type of artificial neural network that is comparable to principal component analysis. The procedure is tested on artificial data simulated from a factor-based population and is compared with existent computational approaches to develop short forms. Autoencoders require mild assumptions on data characteristics and provide a method to predict long-form items' responses from the short form. Indeed, results show that they can help the researcher to develop a short form by automatically selecting a subset of items that better reconstruct the original item's responses and that preserve the internal structure of the long-form.