Appendicitis is among the most frequent reasons for pediatric abdominal surgeries. Previous decision support systems for appendicitis have focused on clinical, laboratory, scoring, and computed tomography data and have ignored abdominal ultrasound, despite its noninvasive nature and widespread availability. In this work, we present interpretable machine learning models for predicting the diagnosis, management and severity of suspected appendicitis using ultrasound images. Our approach utilizes concept bottleneck models (CBM) that facilitate interpretation and interaction with high-level concepts understandable to clinicians. Furthermore, we extend CBMs to prediction problems with multiple views and incomplete concept sets. Our models were trained on a dataset comprising 579 pediatric patients with 1709 ultrasound images accompanied by clinical and laboratory data. Results show that our proposed method enables clinicians to utilize a human-understandable and intervenable predictive model without compromising performance or requiring time-consuming image annotation when deployed. For predicting the diagnosis, the extended multiview CBM attained an AUROC of 0.80 and an AUPR of 0.92, performing comparably to similar black-box neural networks trained and tested on the same dataset.

This practice guideline follows the mission of the World Association of Perinatal Medicine in collaboration with the Perinatal Medicine Foundation, bringing together groups and individuals throughout the world, with the goal of improving the use of antenatal corticosteroids (ACS) for fetal maturation. In fact, this document provides further guidance for healthcare practitioners on the appropriate use of ACS with the aim to increase the timely administration and avoid unnecessary or excessive use. Therefore, it is not intended to establish a legal standard of care. This document is based on consensus among perinatal experts throughout the world and serves as a guideline for use in clinical practice.

Introduction: To investigate the ability of cardiovascular plasma biomarkers to identify imminent preeclampsia (PE) among pregnant women at triage. Material and Methods: C-terminal pro-arginine vasopressin (copeptin), C-terminal pro-endothelin-1 (CT-proET-1), mid-regional pro-adrenomedullin (MR-proADM), and mid-regional pro-atrial natriuretic peptide (MR-proANP) were prospectively measured in pregnant women presenting at the obstetrical triage units of the University Hospitals of Basel and Zurich, Switzerland. Logistic regression and receiver operating characteristics (ROC) analysis was used to assess and quantify the predictive ability of cardiovascular biomarkers. Results: Of the 147 included women, 27 (18.4%) were diagnosed at admission with PE. All biomarker levels were significantly higher in participants with PE as compared to controls. However, only MR-proANP, MR-proADM and CT-proET-1 were significant and independent predictors of PE, after taking into account the effect of various clinical confounders. The area under the ROC curve (AUC) was 0.62 (95% confidence interval 0.50-0.73) for copeptin, 0.64 (0.52-0.76) for MR-proADM, 0.71 (0.61-0.82) for CT-proET-1, and 0.83 (0.73-0.92) for MR-proANP. The combination of MR-proANP and MR-proADM resulted in the highest diagnostic performance (AUC 0.88; 0.79-0.96). Discussion: Assessment of the cardiovascular plasma biomarkers MR-proANP and MR-proADM holds promise to support diagnosis of PE at triage.

Background and purpose: To determine whether neurofilament light chain (NfL), a promising serum and cerebrospinal fluid (CSF) biomarker of neuroaxonal damage, predicts functional outcome in preterm infants with neonatal brain injury. Methods: Our prospective observational study used a sensitive single-molecule array assay to measure serum and CSF NfL concentrations in preterm infants with moderate to severe peri/intraventricular hemorrhage (PIVH). We determined temporal serum and CSF NfL profiles from the initial diagnosis of PIVH until term-equivalent age and their association with clinical and neurodevelopmental outcome until 2 years of age assessed by Bayley Scales of Infant Development (3rd edition). We fitted univariate and multivariate logistic regression models to determine risk factors for poor motor and cognitive development. Results: The study included 48 infants born at < 32 weeks of gestation. Median serum NfL (sNfL) at PIVH diagnosis was 251 pg/mL [interquartile range (IQR) 139–379], decreasing markedly until term-equivalent age to 15.7 pg/mL (IQR 11.1–33.5). CSF NfL was on average 113-fold higher (IQR 40–211) than corresponding sNfL values. Additional cerebral infarction (n = 25)-but not post-hemorrhagic hydrocephalus requiring external ventricular drainage (n = 29) nor any other impairment-was independently associated with sNfL. Multivariate logistic regression models identified sNfL as an independent predictor of poor motor outcome or death at 1 and 2 years. Conclusions: Serum neurofilament light chain dynamics in the first weeks of life predict motor outcome in preterm infants with PIVH.