| 摘要: | 在傳統心臟聽診檢查心血管疾病中(瓣膜狹窄、閉鎖不全)，主要是由醫生個人的經驗和較為主觀的判斷，在許多的診斷中較難討論出一個客觀的標準。因此本研究提出了一款可攜式的數位聽診工具，藉由將心音訊號數位化，完整呈現生理訊號的聲音、另一面將所儲存的生理訊號分析，可藉由機器學習，和資料庫中的類似案例做比對，提供一個客觀輔助診斷。
系統主要可分為類比電路和數位訊號處理這兩個部分。
其中類比電路部分包含兩組不同頻段的二階帶通濾波器(分別是心音訊號 : 20~1k Hz和動靜脈簍管音訊號 : 50~1k Hz)，再結合一顆全頻段的麥克風以及機構設計自製的聽診頭。
當前端收完訊號後，類比數位轉換器把前端的訊號經過微處理器程式編程下在穩定不掉點的狀態下能夠以6.4 kHz取樣頻率做訊號擷取，再利用超低功耗藍芽同步傳輸data到手機端做及時的生理訊號監測，同時把訊號透過音頻放大電路的優化，讓過濾掉雜訊的生理訊號得以在耳機清楚的監聽，最後將raw data儲存在外掛的 Micro SD卡模組，方便在電腦端進行計算與分析，從聽診頭、類比電路設計到微處理器的程式設計，訊噪比可以達到 : 50.6(dB)，使得到的生理訊號能夠作為較客觀並完整的診斷。
本裝置除了能以高解析數位化方式來錄製量測到的生理音，還有手機APP即時監測訊號以及超低功耗、體積小、便於攜帶操作等優點。未來之應用不僅僅只限於大型醫院或加護病房等機構，也可普及於小型醫療機構、戶外急救單位甚至是個人居家，實現即時且普及的生理訊號監測。
;In the traditional auscultation examination for cardiovascular diseases (valvular stenosis, atresia), the doctor′s personal experience and analysis is a diagnostic criterion that is difficult to discuss in this diagnosis. Therefore, the study puts forward its own diagnostic criteria. This is a portable digital auscultation tool. The data consists of digitizing the heart sound signal to fully present the sound of the physical signal. On the other hand, it analyzes the stored rational signal, which can be compared with similar cases in the library by machine learning, and provides Auxiliary diagnosis.
The system can be divided into two parts: analog circuit and digital signal processing.
A second-order band communication with different sounds than the circuit part (respectively the heart sound signal : 20~1k Hz and the rattle tube sound signal: 50~1k Hz), combined with a full series of themes and the organization design itself Auscultation head.
After the front end receives the signal, the analog-to-digital converter converts the front end signal. It can use 6.4 kHz micro signal for signal collection in the state of drop, and then use the ultra-low standby Bluetooth to synchronously transmit data to the mobile phone for timely physiological signal monitoring, and at the same time filter the signal through audio amplification The circuit is optimized. The physiological signals that drop the noise can be monitored on the earphones. Finally, the original data is saved in the external micro SD card module, which is convenient for calculation and analysis on the computer side, from the auscultation head, analog circuit design to program design, the communication establishment ratio can reach: 50.6 (dB), write your own physiological communication number as a larger scale and complete diagnosis.
In addition to being able to capture and measure physiological sounds in a high-resolution digital way, this device has the advantages of real-time signal monitoring via APP, time-consuming, small size, and portable operation. Or institutions such as wards, small medical institutions for patients, out-call rescue units, and even personal home communications, to achieve immediate and unenhanced physiological number monitoring. |
