Search

Background/Objectives: Extrauterine growth restriction (EUGR) is associated with high mortality and an increased incidence of poor neurodevelopmental outcomes in preterm infants. In this study, we aimed to compare the Intergrowth-21ST (IG-21ST) and Fenton charts in predicting long-term neurodevelopmental and anthropometric outcomes of very low birth weight (VLBW) infants. Methods: Data were collected from 2649 VLBW infants registered in the Korean Neonatal Network born between 240/7 and 316/7 weeks of gestational age from January 2013 to December 2017. Follow-up assessments were conducted at 18–24 months of age, corrected for prematurity. Multiple logistic regression analysis was performed to evaluate the association between EUGR and long-term outcomes. Results: Among the 2649 VLBW infants, 60.0% (1606/2649) and 36.9% (977/2649) were diagnosed as having EUGR defined by the Fenton chart (EUGRF) and by the IG-21ST chart (EUGRIG), respectively. The EUGRIG group exhibited a higher proportion of infants with cerebral palsy, neurodevelopmental impairment (NDI), and growth failure. In multiple logistic regression analysis, adjusted for risk factors for long-term outcome, the EUGRIG group showed higher risk of cerebral palsy (adjusted odds ratio [aOR], 1.66; 95% confidence interval [CI], 1.04–2.65), NDI (aOR, 2.09; 95% CI, 1.71–2.55), and growth failure (aOR, 1.57; 95% CI, 1.16–2.13). Infants with EUGRF tended to develop NDI (aOR, 1.29; 95%CI, 1.03–1.63) and experience growth failure (aOR, 2.44; 95% CI, 1.77–3.40). Conclusions: The IG-21ST chart demonstrated a more effective prediction of long-term neurodevelopmental outcomes, whereas the Fenton chart may be more suitable for predicting growth failure at 18–24 months. [ABSTRACT FROM AUTHOR]

Pedestrian detection is a critical task for safety-critical systems, but detecting pedestrians is challenging in low-light and adverse weather conditions. Thermal images can be used to improve robustness by providing complementary information to RGB images. Previous studies have shown that multi-modal feature fusion using convolution operation can be effective, but such methods rely solely on local feature correlations, which can degrade the performance capabilities. To address this issue, we propose an attention-based novel fusion network, referred to as INSANet (INtra-INter Spectral Attention Network), that captures global intra- and inter-information. It consists of intra- and inter-spectral attention blocks that allow the model to learn mutual spectral relationships. Additionally, we identified an imbalance in the multispectral dataset caused by several factors and designed an augmentation strategy that mitigates concentrated distributions and enables the model to learn the diverse locations of pedestrians. Extensive experiments demonstrate the effectiveness of the proposed methods, which achieve state-of-the-art performance on the KAIST dataset and LLVIP dataset. Finally, we conduct a regional performance evaluation to demonstrate the effectiveness of our proposed network in various regions. [ABSTRACT FROM AUTHOR]