女性乳腺癌手掌特征提取及易感性研究

Breast cancer is one of the most frequently diagnosed cancers. Early diagnosis is the key to enhance the living probability of breast cancer. The aetiology of breast cancer is associated with genetic and environmental influence, which has not been completely clarified yet. Therefore, susceptibility identification is difficult. Dermatoglyphy is the gene appeared on the skin, influenced by environmental factors, and is in the nature of high stability and individual identification. According to the latest leading research of genetics, some hereditary diseases are reflected on dermatoglyphy. Moreover, breast and dermatoglyphy are originaed from ectoderm, finger differentiation and breast mutation are both related to the same gene family. Therefore, we speculate that palm may contain specific features of breast cancer based on theoretical analysis. Current research on palm features specific to breast cancer demonstrates several drawbacks such as simplicity, weak specificity and objectivity. Systematic research on palm features related to breast caner and susceptibility detection on early stage have not been reported yet. Therefore, we take breast cancer subjects of female minority groups， whose gene has high single stability, and the associated healthy groups as research subjects. We plan to (1) establish palm image database of subjects, (2) propose feature extraction approaches to obtain the texture, topologic, symmetric and minutiae features of palm, (3) determine sensitive characteristic parameters, study the approach for the auto-classifying the selected characteristic parameters and (4) develop the non-invasive and portable susceptibility detection system on early stage before breast cancernces.The project aims at providing evidences and methods for susceptibility detection and prevention for breast cancer.

乳腺癌是女性发病率最高的恶性肿瘤之一。提高乳腺癌患者生存率的关键在于早期发现和治疗。乳腺癌病因与遗传及环境因素相关，至今未完全阐明，易感预测困难。手掌纹理是人体暴露在外的遗传因子，也受环境因素影响，具有个体特异性和高度的稳定性。遗传学的先导理论认为：人体的遗传性病变会体现在手纹上，乳腺与手纹同源于外胚层而且手指分化及乳腺癌变异与同一基因族相关，因此提示手掌可能包含乳腺癌的特异性特征。现有乳腺癌相关手掌特征研究选取的特征简单、特异性不强、准确性因人而异，基于手掌特征对乳腺癌易感预测的系统性研究还未见报道。项目以基因相对单一稳定的少数民族女性乳腺癌患者和健康对照组为研究对象，建立手掌样本数据库；研究手掌的纹理、拓扑、对称性及纹线细节特征提取算法；筛选高敏感特征，研究特征分类，建立无损便携的易感预测系统。以期为乳腺癌的前瞻性易感预测和防治提供手段和依据。

乳腺癌是女性发病率最高的恶性肿瘤之一。提高乳腺癌患者生存率的关键在于早期发现和治疗。乳腺癌病因与遗传及环境因素相关，至今未完全阐明，易感预测困难。手掌纹理是人体暴露在外的遗传因子，也受环境因素影响，具有个体特异性和高度的稳定性。遗传学的先导理论认为：人体的遗传性病变会体现在手纹上，乳腺与手纹同源于外胚层而且手指分化及乳腺癌变异与同一基因族相关。因此提示手掌可能包含乳腺癌的特异性特征。现有乳腺癌相关手掌特征研究选取的特征简单、特异性不强、准确性因人而异，基于手掌特征对乳腺癌易感预测的系统性研究还未见报道。. 通过本项目的研究，完成了以下内容：（1）以基因相对单一稳定的少数民族女性乳腺癌患者和健康对照组为研究对象，建立手掌样本数据库；（2）研究并实现了手掌的纹理、拓扑、对称性及纹线细节特征提取算法；（3）筛选高敏感特征，研究特征分类，建立无损便携的易感预测系统。论文的研究成果以期为乳腺癌的前瞻性易感预测和防治提供手段和依据。. 在项目研究期内，发表科研论文19篇。申请人第一作者及通信作者发表了科研论文13篇，其中SCI期刊论文2篇，EI期刊论文7篇，EI会议论文1篇，中文核心期刊论文5篇。在国家一级出版社出版教材及专著2部。申请并公开专利3项。获得软件著作权3项。在本项目的支持下，申请人获得了云南省万人计划“教学名师”项目资助。获得Google支持教育部产学合作协同育人项目。在本项目研究期间，通过项目研究反哺教学，申请人获得了云南大学17届教学成果一等奖和2017年云南省省级教学成果二等奖，以主持人获得云南大学校级教改项目资助2项。申请人在项目研究期间，晋升为教授及博导。