Artificial Intelligence Medical Compendium

Explore the latest research on artificial intelligence and machine learning in medicine.

Showing 33,991 to 34,000 of 221,510 articles

Advanced Polymeric Physical Unclonable Functions: From Mechanistic Design and Computer Vision to Enhanced Effect of Stimuli.

Small (Weinheim an der Bergstrasse, Germany)
Commodity fraud is a serious issue leading to significant financial losses for producers, marketers, and consumers. The past anticounterfeiting and security systems were cloneable and predictable. Today, development of advanced, multilayered, and unc... read more 

Innovation in intensive care: a framework to turn ideas and concepts into actionable solutions.

Intensive care medicine
INTRODUCTION: Intensive care is under increasing pressure due to demographic changes, availability of new complex therapies, and workforce shortages, resulting in a relative reduction in resources. Technological advances such as digitalization, weara... read more 

Medical image pretraining-based transfer learning for generalizable and robust diagnosis of bone tumors on radiographs: a multi-center study.

Insights into imaging
OBJECTIVES: To develop a generalizable and robust deep learning model for bone tumor classification in radiographs by leveraging domain-specific medical image pretraining. MATERIALS AND METHODS: This retrospective multi-center study included 2338 pat... read more 

Beyond identification: inferring physical activity from fingerprint lipids using machine learning.

Analytical methods : advancing methods and applications
Fingerprints are a widely recognized form of forensic evidence, valued for their ability to link individuals with specific locations. Traditional fingerprint analysis relies on optical imaging to identify a match in a fingerprint database; however, w... read more 

Patch Uniform Fusion Transformer with circle-inspired optimized walrus-based feature selection for high-dimensional data analysis.

Journal of molecular modeling
PURPOSE: Feature selection approaches have historically been limited by processing performance during high-dimensional data analysis, which presents additional inefficiencies when dealing with bigger datasets. Hybrid search methods are also great and... read more 

Integrated approach for arsenic prediction and health risk evaluation in community tube wells installed by public health department: comparative study of random forest, extreme gradient boosting, and deep neural networks.

Environmental geochemistry and health
Arsenic contamination in groundwater poses a significant public health risk, especially in developing countries with inadequate water quality monitoring systems. This study employs advanced machine learning approaches to assess arsenic pollution in 2... read more 

Automated contrast-to-noise ratio analysis in chest CT: validation of an open-source segmentation approach.

Insights into imaging
OBJECTIVES: This study aimed to evaluate the feasibility and accuracy of automated contrast-to-noise ratio (CNR) analysis in chest CT using the open-source body and organ analysis (BOA) framework and to validate segmentation modifications for reprodu... read more