相关实验视频

Updated: May 20, 2025

A Knowledge Graph Approach to Elucidate the Role of Organellar Pathways in Disease via Biomedical Reports

A Knowledge Graph Approach to Elucidate the Role of Organellar Pathways in Disease via Biomedical Reports

Published on: October 13, 2023

基于知识图的几次学习,用于医学成像报告的标签

Tiancheng Li¹, Yuxuan Zhang², Deyu Su¹

¹The First Affiliated Hospital of Anhui Medical University, Anhui Medical University, Hefei 230032, China (T.L., D.S., J.T.); Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei, China (T.L., D.S., C.L., J.T.).

Academic radiology

|March 27, 2025

概括

这项研究引入了一种新的数据增强方法,使用知识图和少数射击学习来改进人工智能驱动的自动放射学报告标签. 这种方法显著提高了标签准确性和效率,减少了手工工作量.

科学领域:

医疗成像医学成像
人工智能的人工智能
数据科学数据科学数据科学

背景情况:

对人工智能应用程序的大型医学成像数据集的手动标签是一个重要的瓶.
现有的AI模型在自动成像报告标签的规模和复杂性方面扎.

研究的目的:

开发一个有效的数据增强策略,用于在自动放射学报告标签中进行少数拍摄的学习.
利用知识图 (KG) 来增强人工智能模型的数据增强.

主要方法:

构建了腰椎X射线图像的KG,并注释了2000个数据点.
使用KG属性 (同义词/替代) 增强训练数据,并训练了一个BERT模型.
用不同的增强比率和方法评估性能,重点是精度和F1分数.

主要成果:

数据增强显著改善了模型性能,精度达到0.932,F1得分为0.881,比例为1:100.
将同义词和替代增强结合起来,产生了最高的F1得分 (0.881).
对表现不佳节点的代数据补充进一步提高了F1平均得分,达到0.979.

结论:

以知识图驱动的数据增强对于训练AI模型在有限的数据集上进行放射学报告标签是有效的.
这种方法大大减少了手工标签的努力,并提高了标签的准确性和效率.
提出了一个可行的战略,用于推进AI应用在医疗图像报告自动化.

关键词:

有几次射击学习学习.知识图表知识图表自然语言处理自然语言处理.放射学报告放射学报告

更多相关视频

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

相关概念视频

Positron Emission Tomography

Positron Emission Tomography

Positron emission tomography (PET) is a medical imaging technique involving radiopharmaceuticals — substances that emit short-lived radiation. Although the first PET scanner was introduced in 1961, it took 15 more years before radiopharmaceuticals were combined with the technique and revolutionized its potential.
One of the main requirements of a PET scan is a positron-emitting radioisotope, which is produced in a cyclotron and then attached to a substance used by the part of the body...