基于CNN双路特征融合模型的秀丽隐杆线虫寿命阶段分类研究  

Research on Classification of Caenorhabditis Elegans Life Stage Based on CNN Dual-Path Feature Fusion

作　　者：尹彦豪 刘俊[1,2] 杨烨 YIN Yanhao;LIU Jun;YANG Ye(College of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065;Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System,Wuhan 430065;College of Life Science and Technology,Huazhong University of Science and Technology,Wuhan 430074)

机构地区：[1]武汉科技大学计算机科学与技术学院,武汉430065 [2]智能信息处理与实时工业系统湖北省重点实验室,武汉430065 [3]华中科技大学生命科学与技术学院,武汉430074

出　　处：《计算机与数字工程》2023年第5期1163-1168,共6页Computer & Digital Engineering

摘　　要：秀丽隐杆线虫(C.elegans)由于其有着寿命较短和人类基因高度同源的特点,作为一种被优秀的模式生物,被广泛应用于多种人类健康和疾病模型的研究中。为了准确测定当前线虫所处于的寿命阶段,论文提出了一种基于CNN双路特征融合模型的的秀丽隐杆线虫寿命阶段分类方法。首先,利用卷积神经网络分类模型对线虫图像进行特征提取;同时使用快速傅里叶变换、LoG算子等图像处理算法得到荧光蛋白亮斑坐标,然后根据坐标信息计算得到荧光蛋白亮斑聚集度信息;最后,将前两步得到的特征拼接通过分类器得到分类结果。实验结果表明,论文方法能够准确、有效地对线虫寿命阶段进行分类。As an excellent model organism,C.elegans has been widely used in many human health and disease researches due to its short life span and high homology with human genes.In order to accurately determine the life stage of Caenorhabditis elegans,a classification method of caenorhabditis elegans life stage based on CNN Dual-Path feature fusion model is proposed.Firstly,the convolutional neural network classification model is used to extract the features of the nematode image.At the same time,fast Fourier transform,LoG operator and other image processing algorithms are used to obtain the fluorescent protein bright spot coordinates.Then the degree of aggregation of fluorescent protein bright spots is calculated according to the coordinate information.Finally,the features obtained in the first two steps are spliced through the classifier to obtain the classification result.Experimental results show that this method can accurately and effectively classify the life stages of nematodes.

关 键 词：秀丽隐杆线虫 图像分类 深度学习 快速傅里叶变换 拉普拉斯-高斯算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]