【作者】 张明亮；

【导师】 刘航；

【作者基本信息】 大连理工大学 ， 生物医学工程（专业学位）， 2021， 硕士

【摘要】 人体呼出的气体中含有独特的新陈代谢信息,这些新陈代谢信息和某些疾病有着对应关系。随着移动医疗的发展,人体疾病无损化检测成为医学诊断的研究热点。研究人体呼出气体与疾病之间存在的对应关系具有重要的研究意义,其在医疗诊断和监测人体健康等领域具有广泛的应用前景。通过人体呼出气体检测疾病的技术已经广泛应用于医疗检测中,但是目前市面上的呼吸检测仪器普遍存在操作复杂、检测成本高等缺点。本文设计了一款基于云端的呼出气信号检测系统,包括呼出气信号硬件采集系统和软件分析算法。硬件系统可以采集抽烟人群呼出气体中的不同成分信息,软件分析算法可以根据人体呼出气体中不同成分浓度的变化,判断被检测人在检测前持续未抽烟的时间。目前该系统可以应用于日常居家中戒烟人群的监测。本文的主要工作如下:(1)设计并搭建了呼出气信号硬件采集系统。根据系统硬件部分元器件的功能、工作环境,遵循系统的设计方案、技术指标等,在Altium Designer 19中画出了数据采集电路原理图,并根据原理图完成了PCB板的布线和绘制。通过蓝牙将数据传输到手机端.完成硬件系统的设计与实物搭建。在Arduino编程软件中设计了呼出气信号的输出模式、传输速率等。在Android Studio中设计了手机端应用程序,将手机端保存的传感器阵列数据传输到配置好的阿里云服务器中存储。(2)设计了呼出气信号分析算法。本文设计的硬件系统主要通过传感器阵列采集数据,系统采集到的呼出气信号是一维时间序列数据,通过格拉姆求和角场(Gramian Angular Summation Field,GASF)进行数据处理,将一维时间序列数据转换成二维图谱。利用ResNet18神经网络进行图谱识别,完成对目标气体的检测。(3)进行大量的数据采集实验。对抽烟人群在持续未抽烟的不同时间段的呼出气信号进行采集(刚抽完烟、抽完烟1小时、抽完烟2小时、抽完烟4小时以及早上没有抽烟五个阶段),系统一共采集了800人次的有效数据。将采集到的数据进行处理,利用设计的识别算法完成对抽烟人最近一次未抽烟持续时间的识别。本文设计的呼出气体检测系统,实现了从最初的气体信息采集、数据传输、数据处理,到最后对目标气体进行识别。经过数据采集和识别实验验证表明,系统各个功能模块达到了设计要求。系统体积小、检测成本低、能够实时采集数据,可以作为医疗检测的辅助工具在日常居家中使用。更多还原

【Abstract】 The breath exhaled by the human body contains unique metabolic information,which has a corresponding relationship with certain diseases.With the development of mobile medicine,non-destructive detection of human disease has become a hotspot research issue in medical diagnosis.The research on the corresponding relationship between human exhaled gas and disease is of great significance,and it has broad application prospects in the fields of medical diagnosis and human health monitor.The technology of detecting diseases through human exhaled gas has been widely used in medical testing,but the current respiratory testing instruments on the market generally have disadvantages such as complicated operation and high testing cost.This paper designed a cloud-based exhalation signal detection system,including exhalation signal hardware acquisition system and software analysis algorithm.The hardware system can collect the information of different components in the exhaled air of the smoker,and the software analysis algorithm can determine the time that the detected person has not smoked before the detection according to the changes in the concentration of different components in the exhaled air of the human body.At present,the system can be applied to the monitoring of people who quit smoking at home.The main work of this paper is as follows:(1)Designed and built a hardware acquisition system for exhaled breath signal.According to the functions and working environment of the system hardware components,following the system design plan,technical indicators,etc.,the data acquisition circuit schematic diagram was drawn in Altium Designer 19,and the PCB board wiring and drawing were completed according to the schematic diagram.The data is transmitted to the mobile phone via Bluetooth.Completed the design and physical construction of the hardware system.The output mode and transmission rate of the exhalation signal were designed in the Arduino programming software.A mobile phone application was designed in Android Studio to transfer the sensor array data saved on the mobile phone to the configured Alibaba Cloud server for storage.(2)Designed the exhalation signal analysis algorithm.The hardware system designed in this paper collects data mainly through sensor arrays.The exhaled breath signal collected by the system is one-dimensional time series data.The data is processed through Gramian Angular Summation Field(GASF)map technology to convert the one-dimensional time series data into a two-dimensional map.Using ResNet18 neural network to perform atlas recognition to complete the detection of target gas.(3)Conducted a large number of data collection experiments.Collect the exhalation signals of smokers during different periods of time without smoking(5 stages after smoking,1hour after smoking,2 hours after smoking,4 hours after smoking,and no smoking in the morning),the system a total of 800 valid data were collected.The collected data was processed,and the designed recognition algorithm was used to complete the recognition of the last non-smoking duration of the smoker.The exhaled gas detection system designed in this paper realized the initial gas information collection,data transmission,data processing,and finally the identification of the target gas.The data collection and identification experiments showed that the various functional modules of the system meet the design requirements.The system is small in size,low in detection cost,and can collect data in real time.It can be used as an auxiliary tool for medical detection in daily home.更多还原