Search research articles

关于 JoVE

概览领导团队博客 JoVE 帮助中心

作者

出版流程编辑委员会范围与政策同行评审常见问题投稿

图书馆员

用户评价订阅访问资源图书馆顾问委员会常见问题

研究

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments 存档

教育

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual 教师资源中心教师网站

使用条款与条件

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序

Same author

Differences in clozapine metabolism across geographical regions in Turkiye.

Dusunen adam : Bakirkoy Ruh ve Sinir Hastaliklari Hastanesi yayin organi·2026

Same author

Model-informed precision dosing in paediatric intensive care: Bridging the gap between pharmacological optimization and clinical benefit.

British journal of clinical pharmacology·2026

Same author

Drug-associated acute pancreatitis in paediatric versus adult reports: a disproportionality analysis using FAERS.

European journal of pediatrics·2026

Same author

Time-dependent pattern of liver injury biomarkers in neonates with hypoxic-ischemic encephalopathy undergoing therapeutic hypothermia.

European journal of pediatrics·2026

Same author

Food Origin Authenticity Using Deep Learning and Citizen Science: Bananas Case Study.

Foods (Basel, Switzerland)·2026

Same author

Extracorporeal circuit-related hemodilution in pediatric CKRT: the dilution factor concept.

Pediatric nephrology (Berlin, Germany)·2026

Search research articles

相关实验视频

Updated: May 12, 2025

Deep Neural Networks for Image-Based Dietary Assessment

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

预测NICU患者营养不良的机器学习模型:一项全面的基准研究.

Sander M W Janssen¹, Yamine Bouzembrak², Nadir Yalcin³

¹Information Technology Group, Wageningen University and Research, Wageningen, the Netherlands; Department of Primary and Community Care, Radboud Institute for Health Sciences, Radboudumc, Nijmegen, the Netherlands.

Computers in biology and medicine

|May 8, 2025

概括

机器学习模型为营养不良查提供了有效的替代方案. 使用拉索或弹性网调节 (GLMnet) 和极端梯度提升 (XGBoost) 的通用线性模型显示了营养评估的有希望的结果.

关键词:

支持决定的决定支持.机器学习是机器学习.营养不良营养不良营养评估营养评估营养查工具是一种营养查工具.精准营养是一种精确的营养.

更多相关视频

Assessment and Evaluation of the High Risk Neonate: The NICU Network Neurobehavioral Scale

Assessment and Evaluation of the High Risk Neonate: The NICU Network Neurobehavioral Scale

Published on: August 25, 2014

Determining Gender-Based Differences in Retinal and Choroidal Thickness in Underweight Individuals via Swept-Source Optical Coherence Tomography

Determining Gender-Based Differences in Retinal and Choroidal Thickness in Underweight Individuals via Swept-Source Optical Coherence Tomography

Published on: December 1, 2023

相关实验视频

Last Updated: May 12, 2025

Deep Neural Networks for Image-Based Dietary Assessment

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Assessment and Evaluation of the High Risk Neonate: The NICU Network Neurobehavioral Scale

Assessment and Evaluation of the High Risk Neonate: The NICU Network Neurobehavioral Scale

Published on: August 25, 2014

Determining Gender-Based Differences in Retinal and Choroidal Thickness in Underweight Individuals via Swept-Source Optical Coherence Tomography

Determining Gender-Based Differences in Retinal and Choroidal Thickness in Underweight Individuals via Swept-Source Optical Coherence Tomography

Published on: December 1, 2023

科学领域:

营养科学营养科学
计算生物学计算生物学
医疗信息学医疗信息学

背景情况:

营养不良是影响成人和儿童的全球健康问题,源于营养摄入不足或体重减轻.
传统的营养不良查方法往往资源密集,耗时,缺乏一致的准确性和广泛的适用性.
自动机器学习 (ML) 方法为高效和可适应的营养评估提供了潜在的解决方案.

研究的目的:

评估各种机器学习模型在预测营养不良方面的有效性.
在统一的营养不良数据集上对22种不同的回归和分类模型进行比较.
通过确定最小的输入特征要求来优化模型效率.

主要方法:

为了可重复性,实施了一个强大的模型开发管道,利用新生儿重症监护室 (NICU) 患者数据集 (412名患者,232名用于培训).
测试了广泛的ML模型,包括线性模型,基于树的模型,神经网络和整体方法.
模型对回归和分类任务进行了评估,以预测营养状况.

主要成果:

使用拉索或弹性网调节 (GLMnet) 的通用线性模型实现了回归任务的最高性能,其R平方 (R2) 为0.79.
极端梯度增强 (XGBoost) 模型在分类任务中表现出卓越的性能,其曲线下面面积 (AUC) 为0.79.
无论是GLMnet还是XGBoost,都提供了可靠的自动营养评估功能.

结论:

机器学习模型为传统营养不良查工具提供了有效和自动化的替代方案.
该研究成功地确定了用于营养评估的高性能ML模型 (回归的GLMnet,分类的XGBoost).
这些自动化方法可以通过提供高效和准确的营养评估来减少医疗保健系统的负担.