設計及製作可攜式聲學系統於分析肺部生理訊號

在傳統的胸部聽診診斷當中，無論是心音診斷、或是肺音診斷，都相當取決於臨床醫生的經驗和判斷，對於在分別不同的症狀時並沒有一個客觀且可以量化的標準，為了減少聽診的主觀不確定性，透過將聽診時的訊號儲存並影像化，更可以透過機器學習分類具有相同特徵的病人，本研究提出了一款可攜式的電子聽診器，可用於紀錄、儲存和分析心臟和肺部等生理訊號，採用人工智能、改進培訓聽診等方法、將聽診狀況系統化，輔助臨床醫生正確的識別患者的情況並給出適合的幫助。本系統架構由三大部分組成，第一部分是數位控制電路，由類比數位轉換器(Analog-to-Digital Convertor, ADC)、微控制器單元(Microcontroller unit, MCU)及電源管理(Power management)組成。第二部分是量測生理訊號的類比電路，由帶通濾波器濾除非生理訊號頻帶內的雜訊，配合一顆全指向性的電容式麥克風和自製拾音頭，再利用序列周邊介面(Serial Peripheral Interface Bus, SPI)與外掛安全數位卡模組(Secure Digital Memory Card, SD Card)以6.4kHz的取樣率儲存原始訊號，將資料傳送到電腦後，搭配Python、Matlab做解碼、計算和分析等處理。第三部分是手機APP，在聽診的同時，透過通用非同步收發傳輸器(Universal Asynchronous Receiver/Transmitter, UART)及藍芽(Bluetooth Low Energy, BLE)，將訊號即時同步顯示到手機端。由於手機端的即時監測帶來的許多優點，超低功耗、體積較小、便於攜帶操作等優點，未來再應用層面可普及到小型醫療機構或是居家檢測，在目前疫情嚴峻的時期，更可利用雲端的功能實現遠端醫療。

關鍵字

APP即時監測；生理音；聽診

並列摘要

In the traditional chest auscultation diagnosis, whether heart sound diagnosis or lung sound diagnosis depends on the experience and judgment of clinicians. There is no objective and quantifiable standard for different symptoms. In order to reduce the subjective uncertainty of auscultation, the signals during auscultation are stored and imaged, We can also classify patients with the same characteristics through machine learning. This study proposes a portable electronic stethoscope, which can be used to record, store and analyze physiological signals such as heart and lung. We can systematize the auscultation status by using artificial intelligence and improving training auscultation, so as to assist clinicians to correctly identify the situation of patients and give appropriate help. The system architecture consists of three parts. The first part is the digital control circuit, which is composed of analog-to-digital converter (ADC), microcontroller unit (MCU) and power management. The second part is the analog circuit for measuring physiological signals. The band-pass filter filters out the noise in the frequency band of non physiological signals, cooperates with an omnidirectional capacitive microphone and a self-made pickup, and then uses the serial peripheral interface bus (SPI) and the external secure digital memory card (SD card) to store the original signals at a sampling rate of 6.4khz, After the data is transmitted to the computer, it is decoded, calculated and analyzed with Python and MATLAB. The third part is the mobile phone app. While auscultating, the signals are synchronously displayed to the mobile phone through universal asynchronous receiver / transmitter (UART) and Bluetooth low energy (ble). Due to the many advantages brought by the real-time monitoring of the mobile terminal, such as ultra-low power consumption, small size and easy to carry operation, the re application level can be popularized to small medical institutions or home detection in the future. In the current severe epidemic period, the function of the cloud can be used to realize remote medical treatment.

並列關鍵字

APP real-time monitoring ； Physiological sound ； Auscultation

參考文獻

Google Scholar

[1] Teng Tong, Lishen Qiu, Jun Zhong, Chongsen Zang, Lirong Wang, Bin Chen. “Wearable arteriovenous fistula murmur monitoring system based on embedded Wi-Fi technology”. DOI : 10.1109/SmartWorld.2018.00058

Google Scholar

[2] J. E. Earis, B. M .G Cgeetham, “Current methods used for computerized respiratory sound analysis” European Respiratory review, pp.586-590,2000

Google Scholar

[3] Daisuke Higashi, Keisuke Nishijima, Ken’ichi Furuya, Keiko Tanakay, Satoko Shiny. “Classification of Shunt Murmurs for Diagnosis of Arteriovenous Fistula Stenosis”.DOI：10.23919/APSIPA.2018.8659641

Google Scholar

[4] J. C. Chien, M. C. Huang, Y. D. Lin, F. C. Chong, “A study of heart sound and lung sound separation by independent component analysis technique”

Google Scholar

國際替代計量

設計及製作可攜式聲學系統於分析肺部生理訊號

全文下載

主題瀏覽