Results 171 to 180 of about 8,613 (235)

VDLIN: A Deep Learning‐Based Platform for Methylcobalamin‐Inspired Immunomodulatory Compound Screening

open access: yesAdvanced Science, EarlyView.
Using the convolutional neural network model VDLIN, Co7 is identified as a promising therapeutic candidate. Co7 demonstrates distinct advantages over MCB by effectively balancing anti‐inflammatory and immune‐stimulatory functions, making it a potential novel approach for immune modulation.
Xuefei Guo   +6 more
wiley   +1 more source

Reliable AI Platform for Monitoring BCI Caused Brain Injury and Providing Real‐Time Protection

open access: yesAdvanced Science, EarlyView.
BrainGuard enables real‐time and interpretable assessment of brain injury caused by brain computer interface (BCI). Using feature‐based Gaussian process (GP) emulators trained on limited biomechanical data, it efficiently predicts full‐field strain and constructs patient‐specific digital brain twins to support clinical diagnosis and long‐term BCI ...
Chufan He   +3 more
wiley   +1 more source

Continuous Physics‐Informed Learning Expedited Battery Mechanism Decoupling

open access: yesAdvanced Science, EarlyView.
A battery modeling framework combining dual neural networks with physical prior is presented to track internal mechanisms, decouple overpotential components, and early predict aging trajectories using only DC data, eliminating the need for additional measurements.
Shanling Ji   +7 more
wiley   +1 more source

Strong Proton‐Phonon Coupling Drives Fast Ion Transport in Perovskites

open access: yesAdvanced Science, EarlyView.
Experimental and computational phonon analysis of ABO3‐type proton conductor BaSnO3 shows that substitution on the B‐site with yttrium forms an imaginary phonon mode which is instrumental for the function as proton conductor. This overcompensates the adverse proton trapping effect of the yttrium.
Alexey Rulev   +8 more
wiley   +1 more source

Observing hidden neuronal states in experiments. [PDF]

open access: yesPLoS Comput Biol
Amakhin D   +5 more
europepmc   +1 more source

Data‐Driven Printability Modeling of Hydrogels for Precise Direct Ink Writing Based on Rheological Properties

open access: yesAdvanced Science, EarlyView.
A predictive model for 3D printability is developed by integrating rheological analysis, including the Large Amplitude Oscillatory Shear (LAOS) test, with machine learning. With prediction errors under 10%, the model shows that post‐extrusion recovery controls horizontal printability, while high‐strain‐rate nozzle flow dictates vertical printability ...
Eun Hui Jeong   +7 more
wiley   +1 more source

Home - About - Disclaimer - Privacy