基于聚类的红外热图代谢综合征辅助诊断算法研究

Research on cluster and infrared thermography-based diagnosis algorithm for metabolic syndrome

目的研究基于人体红外热图的代谢综合征智能辅助诊断算法。方法通过分析人体红外热图的温度分布特征实现对代谢综合征的智能诊断。首先,进行特征提取,计算红外热图样本中每行温度和每列温度的平均值,获取纵向和横向的特征向量;然后,根据横、纵向特征向量的模值将人体红外热图映射为二维空间中的样本点,并对其进行聚类,构建领域圆函数;最后,根据样本点与领域圆的位置关系实现对代谢综合征(metabolic syndrome,MS)的诊断。结果对70例人体红外热图数据进行训练,其准确率为97%;对25例人体红外热图数据进行测试,诊断结果的准确率为96%。结论针对现有的红外热图代谢综合征诊断方法主要依赖临床医生的主观分析和肉眼诊断,缺乏一定客观性的问题,提出了一种基于聚类的红外热图代谢综合征辅助诊断算法,实现了对代谢综合征准确高效的客观诊断,对重大疾病的防治和社会发展具有重大意义。

Objective Researching an intelligent assistive diagnostic method for metabolic syndrome based on human infrared thermography.Methods The intelligent diagnosis of metabolic syndrome is achieved by analyzing the temperature distribution characteristics of human infrared thermography.Firsdy,feature extraction is performed by calculating the average temperature for each row and column in the infrared thermographic samples,obtaining vertical and horizontal feature vectors.Then,the human body infrared thermographic images is mapped into sample points in a Iwo-dimensional space based on the magnitude of the horizontal and vertical feature vectors,clustered,and domain circles are constructed.Finally,metabolic syndrome is diagnosed based on the relationship between the sample points and the domain circle positions.Results Seventy samples of human body infrared thermographic data was used for training in this study.An accuracy rate of 97%is achieved.Subsequently,25 samples of human body infrared thermographic data were used for testing,and a diagnostic accuracy rate of 96%is obtained.Conclusions Because current diagnostic methods for metabolic syndrome based on infrared thermographic data heavily depend on the subjective analysis and visual inspection by clinical physicians,thereby lack a certain level of objectivity,we introduce an auxiliary diagnostic algorithm for metabolic syndrome based on clustering infrared thermographic data.This algorithm facilitates an objective,accurate,and efficient diagnosis of the metabolic syndrome,which is of great significance to the prevention and treatment of severe diseases and societal development.