Results 281 to 290 of about 2,504,885 (377)

Multi‐Disease Detection in Retinal Imaging Using VNet with Image Processing Methods for Data Generation

open access: yesAdvanced Intelligent Systems, EarlyView.
This study introduces a data augmentation method that expands an ophthalmology dataset by 12x, enhancing robustness and reducing overfitting. A novel VNet architecture improves accuracy by 10% over the original dataset and 5% over Grand Challenge benchmarks.
Samad Azimi Abriz   +3 more
wiley   +1 more source

ChatMolData: A Multimodal Agent for Automatic Molecular Data Processing

open access: yesAdvanced Intelligent Systems, EarlyView.
While large language models (LLMs) struggle with molecular data due to single‐modality limitations, ChatMolData—a multimodal agent for processing databases, images, structure files, and documents—is presented. It combines LLMs with tools for retrieval, structuring, prediction, visualization, and search, achieving > 90% accuracy across 128 tasks.
Yi Yu   +5 more
wiley   +1 more source

Mainstream Artificial Intelligence Technologies in Contemporary Ophthalmology

open access: yesAdvanced Intelligent Systems, EarlyView.
This review explores the latest artificial intelligence (AI) technologies in ophthalmology, focusing on four key data types: medical imaging, electronic health records, robotic‐assisted surgery, and genomics. It examines the structural features, use cases, clinical goals, and evaluation metrics of various AI algorithms, while also introducing emerging ...
Shiqi Yin   +9 more
wiley   +1 more source

Hierarchical Superposition Framework Reveals the Complex Effects of Natural Medicine Formulas

open access: yesAdvanced Intelligent Systems, EarlyView.
A novel hierarchical superposition pharmacological model incorporates the principle of hierarchical structures from physics to simulate the spatiotemporal dynamics of drug combinations, to elucidate the universal law underlying drug combination effects. By modeling cross‐level causal transmission and attenuation, it advances beyond traditional additive
Weifeng Liang   +3 more
wiley   +1 more source

Adapting Image‐Based Models for 1D Data via Spider Plot Transformation and Transfer Learning

open access: yesAdvanced Intelligent Systems, EarlyView.
A novel method enables the use of pretrained image‐based neural networks for complex 1D data, including Raman and mid‐infrared spectra, electrocardiograms, and mass spectrometry. 2D spider plots with false‐color fill enable transfer lerning, therefore enhancing data augmentation and model explainability across diverse spectral and time series datasets.
Azadeh Mokari   +2 more
wiley   +1 more source

Home - About - Disclaimer - Privacy