Hongyin Li

Abstract The prognostic value of body composition phenotypes in coronary artery disease (CAD) remains underexplored. We retrospectively analyzed 178 CAD inpatients who underwent bioelectrical impedance analysis (InBody 770). K-means clustering using skeletal muscle index (SMI) and body fat percentage (PBF) as primary discriminators identified three phenotypes: Sarcopenic (lowest SMI, ~5.3 kg/m²), Robust (highest SMI, ~7.6 kg/m²), and Sarcopenic-Obesity (low SMI ~ 5.8 kg/m² and high PBF ~ 36.9%).

A progressive decay Mixup data augmentation for long-tailed image classification 1School of Engineering, Guangzhou College of Technology and Business, Guangzhou, China 2School of Electronics and Communication Engineering, Guangzhou University, Guangzhou, China DOI 10.7717/peerj-cs.3261 Academic Editor Bilal Alatas Subject Areas Algorithms and Analysis of Algorithms, Artificial Intelligence, Computer Vision, Optimization Theory and Computation, Neural Networks Keywords Long-tailed...

Abstract We present a compact, monolithic optical reference for the frequency stabilized laser of future inter-satellite laser interferometer missions. A prototype based on the integration of a high-finesse cavity and associate optics has been designed to be space compatible while maintaining sufficient stability. The prototype has then been developed with a space-qualified bonding technique, and an in situ multi-degree-of-freedom alignment method.