【作者】 胡彬；

【导师】 任长明； 任超世；

【作者基本信息】 天津大学 ， 软件工程， 2004， 硕士

【摘要】 测量人体成分可以得到正常发育、生长、成熟以及老化进程的数据,提供人体成分正常值范围,有助于人们研究相关疾病的发展趋势,监测疾病的治疗效果。生物电阻抗法(BIA)测量人体成分是近年来广泛应用于人体成分测量领域的新方法。它具有操作简便,无痛无创,费用低廉等优点。因此,充分利用BIA这一简便而相对廉价的测量方法,研制高测量精度的人体成分测量仪在医学领域具有十分重要的意义。运用人工神经网络技术提高现行的BIA人体成分测量仪测量精度是本文提出的一个新思路。本文首先分析了当今流行的人体成分测量仪的工作原理,进而讨论了该仪器数据处理模型上存在的不足。针对此不足,在不改变原有测量方法基本框架的前提下,提出了基于人工神经网络技术的新数据处理模型,并对人体成分测量仪进行了软件系统设计和数据仿真验证。根据针对仿真结果以及包含20个观测值的小样本的分析表明,使用人工神经网络作为改进现有人体成分测量仪测量精度的思想是可行的。在本文研究的基础上,通过增加测量样本的数量,最终可以得到一种既不会增加现有仪器生产成本、又可以大幅度提高测量精度的新型实用的人体成分测量仪。本文的研究具有较高的创新性和实用性,具有较好的应用前景。更多还原

【Abstract】 Information can be given by measuring human body com-position for natural growth, maturing and aging with providingthe right value extent of body composition. It is helpful toresearch the developing trend of correlative diseases, andinspect their curative effects. Method of bioelectrical im-pedance analysis (BIA) for human body composition measurementis the most popularly used in this field. It has the ad-vantage of convenient operating, no hurting and cheap charging.It is beneficial using good precision body measurement equip-ment in such domains as families, community hospitals, gym-nasiums and so on.In this these, without changing BIA method, a new proto-type of the body composition measurement equipment based onANN is put forward.This thesis introduces the design and implementation ofa prototype of the body fat mass measurement equipment basedon ANN. At the beginning of the theses, the methods, prin-ciples and significance for body measurement are introduced.as traditional human body measurement method, BIA is detailedintroduced for its working principle. Next, how does a bodyfat mass measurement equipment applying BIA work is introduced.Then , the disadvantage about its data processing model isdiscussed. Aiming at the matter of the pattern, withoutchanging applying BIA method , a new data processing modelbased on ANN is put forward as a possible solution. And then,the hardware implementation is introduced. The details ofthe software design and its implementation are made after that.Subsequently, experiments are made by system’s emulation ofsamples. The assessment about the emulation is made followedthat. In the end, combining the engineering application andthe system’s design, some details of the application ANN in domainof measuring body composition are discussed and the followingresearch work is brought forward.The research in this thesis has the quality of being originaland practicable. It is believed to be in good application.更多还原