221名男性皮肤图像颜色纹理特征与年龄的相关性  

Correlation between color,texture characteristic and age of skin images in 221 men

作　　者：陈展鸿 江涛 许家佗[1] 马旭翔 周锐 王意钦 刘昊华 张斯愉 CHEN Zhanhong;JIANG Tao;XU Jiatuo;MA Xuxiang;ZHOU Rui;WANG Yiqin;LIU Haohua;ZHANG Siyu(School of Traditional Chinese Medicine,Shanghai University of Traditional Chinese Medicine,Shanghai 201203,China;School of Integrative Medicine,Shanghai University of Traditional Chinese Medicine,Shanghai 201203,China;Clinical Medical College of Traditional Chinese Medicine,Shanghai Municipal Hospital of Traditional Chinese Medicine,Shanghai University of Traditional Chinese Medicine,Shanghai 200071,China;School of Acupuncture-Moxibustion and Tuina,Shanghai University of Traditional Chinese Medicine,Shanghai 201203,China;Clinical Medical College,Longhua Hospital Shanghai University of Traditional Chinese Medicine,Shanghai 200032,China)

机构地区：[1]上海中医药大学中医学院,上海201203 [2]上海中医药大学中西医结合学院,上海201203 [3]上海中医药大学附属市中医医院临床医学院,上海200071 [4]上海中医药大学针灸推拿学院,上海201203 [5]上海中医药大学附属龙华医院临床医学院,上海200032

出　　处：《中华中医药杂志》2024年第12期6821-6825,共5页China Journal of Traditional Chinese Medicine and Pharmacy

基　　金：国家自然科学基金项目(No.82104738);国家重点研发计划中医药现代化研究重点专项(No.2017YFC1703301);中国博士后科学基金面上项目(No.2023M732337);上海市科学技术委员会地方院校能力建设(No.21010504400);上海市“超级博士后”激励计划(No.2022509)。

摘　　要：目的:研究男性的中医望诊皮肤图像特征与年龄、季节变化的关系。方法:选用标准化便携式望诊终端Mi-10采集221名健康男性受试者手臂皮肤图像,提取图像R、G、B、L^(*)、a^(*)、b^(*)、C^(*)_(ab)、h_(ab)、个体类型角(ITA)等颜色特征,采用共生灰度矩阵提取图像第二角力矩(ASM)、对比度(CON)、熵(ENT)、逆差矩(IDM)等纹理特征,分析不同季节、性别的皮肤图像特征差异,以及年龄与颜色、纹理指标的相关性。结果:与冬季组比较,夏季皮肤R、G、L^(*)、a^(*)、b^(*)、C^(*)ab、ASM、CON显著升高(P<0.01,P<0.05),ENT、IDM值显著降低(P<0.05,P<0.01)。年龄与R、G、L^(*)、a^(*)、b^(*)、C^(*)_(ab)、ASM、IDM均呈显著性正相关(P<0.01),与CON、ENT呈显著性负相关(P<0.01)。多元线性回归结果显示,季节、C^(*)_(ab)、CON、ENT均与年龄独立相关(P<0.05)。结论:皮肤图像特征与年龄具有相关性,随着年龄的增长,皮肤颜色加深,纹理变粗糙。Objective:To study the relationship between male skin image features of traditional Chinese medicine(TCM)inspection and age and seasonal changes.Methods:The standardized portable inspection terminal Mi-10 was used to collect 221 skin images of arm skin in healthy male subject.Then we used the software to extract R,G,B,L^(*),a^(*),b,C^(*)_(ab),h_(ab)and ITA color features of the image,and used the symbiotic gray matrix to extract the texture features including ASM,CON,ENT,IDM of the image.Finally,we analyzed the skin image feature differences in different seasons and genders and the correlation between age and color,texture indicators.Results:Compared with the winter,the features R,G,L^(*),a^(*),b^(*),C^(*)_(ab),ASM and CON values of the summer increased(P<0.01,P<0.05),while the ENT and IDM value decreased(P<0.05,P<0.01).Age was positively correlated with R,G,L^(*),a^(*),b^(*),C^(*)_(ab),ASM and IDM(P<0.01),and negatively correlated with CON and ENT(P<0.01).Multiple linear regression results showed that season,C^(*)ab,CON,and ENT were independently correlated with age(P<0.05).Conclusion:Skin image features correlate.with age,skin become darker and texture become rougher.

关 键 词：皮肤数字图像 望诊 年龄 颜色 纹理 相关性 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]