基于神经网络的医学图像特征提取算法研究
Application Research on Medical Image Feature Extraction Based on Neural Network
曾叶纯 1汪辉进1
作者信息
- 1. 池州职业技术学院生物与健康系,安徽池州 247000
- 折叠
摘要
全卷积神经网络(Fully Convolutional Networks,FCN)是一种具有深度结构的神经网络,它在复杂人工智能,尤其是图像识别以及自然语言处理问题中应用广泛.全卷积神经网络应用在医学领域,能为人类健康提供更可靠的数据,文章提出一种基于改进的Adam深度学习算法来完成人体重要器官肝脏影像资料的提取,通过几百张肝脏图片进行训练比对,帮助提高肝脏影像图像的异常部分定位及精准性.
Abstract
Fully Convolutional Networks(FCN)is a kind of neural network with depth structure,which is widely used in solving complex artificial intelligence problems,especially in image recognition and natural language processing.In the medical field,FCN can provide more reliable data for human health.This paper proposes an improved Adam deep learning algorithm to realize the extraction of image data of human livers;and by training and comparing hundreds images of human livers,we can see that it is helpful to improve the localization and accuracy of abnormal liver images.
关键词
神经网络/医学图像/特征提取/应用Key words
neural network/medical images/feature extraction/application引用本文复制引用
基金项目
安徽省高等学校自然科学研究重点项目(2020)(KJ2020A1008)
安徽省优秀人才访学项目(GXGNFX2020162)
安徽省质量工程项目(2022cxtd100)
安徽省质量工程项目(2021jyxm1031)
出版年
2024
NETL
NSTL
万方数据