吞嚥感應貼布黏貼方式之開發與驗證及其於重複唾液吞嚥測試之演算法開發

Abstract

本研究的目標為比較吞嚥感應貼布黏貼方式，以及驗證及其於重複唾液吞嚥測試之應用及演算法開發，此吞嚥感應貼布以同時量測舌骨及甲狀軟骨訊號來做為監測吞嚥訊號的一種非侵入式穿戴裝置，以做為吞嚥評估之標準。本研究亦建立一套資料擷取系統，可同時分析舌骨及甲狀軟骨之特徵波形，用以探討所量測到的訊號與吞嚥之間的關係。實驗結果顯示無論使用橫式或直式吞嚥感應貼布，皆可成功擷取到受試者的吞嚥行為，並且比較出健康受試者與吞嚥困難受試者的差異。本研究所開發的演算法對於重複唾液吞嚥測試之準確率在健康男性、健康女性、及患有吞嚥困難受試者中吞嚥次數監測能力在誤差+/-1以內皆可達100%。本研究亦使用超音波影像分析舌骨及甲狀軟骨的運動行為，用以驗證壓電吞嚥感測貼布訊號的可靠度，其中舌骨的總行程時間之誤差範圍達2.04~3.31%，甲狀軟骨的總行程時間之誤差範圍達1.06~4.24%，成功驗證壓吞嚥感應貼布所量測到的訊號可代表舌骨及甲狀軟骨進行吞嚥行為時所產生的喉部皮膚表面形變。總結，本研究以實際實驗驗證吞嚥感應貼布可以建立人體吞嚥行為參數以及準確判斷吞嚥次數，並且在吞嚥困難的受試者身上得到初步試驗與驗證，以及驗證壓電吞嚥感測貼布訊號的可靠度，可作為即時吞嚥監測之個人化裝置。

The aim of this study is to compare the development and validation of different attachment methods for the swallow-patch-sensor and to develop an algorithm for the application of the repetitive saliva swallowing test. The swallow patch sensor measures the skin movement of hyoid bone and thyroid cartilage during the swallowing process for assessment. A data acquisition system is created to simultaneously analyze the characteristic waveforms of the thyroid cartilage and hyoid bone signals to monitor the swallowing process. The experimental results show that the swallowing behavior can be successfully captured using both the horizontal or vertical type swallow patch sensor, and can both identify differences between healthy participants and dysphagia patients. The accuracy of the developed algorithm can reach 100% for healthy male, healthy female and dysphagia subjects with +/-1 error in the total number of swallows, respectively. Finally, ultrasonic images are also used to analyze the behavior of hyoid bone and thyroid cartilage to verify the reliability of the swallow patch sensor. The error of the total travel time of the hyoid bone is 2.04% to 3.31%, and the error of the total travel time of the thyroid cartilage is 1.06% to 4.24%. It verified that the signals measured by the swallow patch sensor can represent the surface deformation of the skin caused by the hyoid bone and thyroid cartilage during the swallowing process. In summary, this study experimentally verifies that using the swallow patch sensor can assess human swallowing behaviors. It can accurately determine the number of swallows in the repetitive saliva swallowing test. It also provides initial experimental verifications on dysphagia participants. Finally, the reliability of this swallow monitoring system is verified, which can become as a personalized device for real-time swallowing monitoring.