用Leap Motion實作手指顫抖的分析並與帕金森氏症做連結

Abstract

Tremogram是目前醫學臨床應用來偵測手指顫抖的儀器，其利用加速規作為感測器來進行手指顫抖的偵測，並且利用輔助軟體偵測結果進行快速傅立葉轉換來得知手指顫抖的頻率，將其作為其中一項判斷帕金森氏症患者嚴重程度的依據。 本論文目的為在Leap Motion上面設計演算法來同時進行五根手指三個軸向的手指顫抖偵測與分析，希望藉由本論文的結果可以幫助臨床醫師對於帕金森氏症患者做出更好的診斷與評估。論文中總共分析了兩組介於20-24歲的正常年輕人以及兩組帕金森氏症的患者，從實驗結果可以看出，對於病患的手指顫抖頻率偵測的結果很接近目前臨床上所使用的Tremogram，更加進一步地提出了手指顫抖的量化以及三維空間的分析。本論文成功地利用成本較低的機器分析出比Tremogram更為多樣化的診斷依據，期待本論文所實作的系統未來可以進一步地應用在臨床醫學上面。

Tremogram is the instrument which is clinically used for measuring fingers trembling. Tremogram is equipped with the accelerometer as the sensor to detect the fingers trembling and then it collocates with the software which can handle the Fast Fourier Transform to get the frequency of fingers trembling. The frequency is one of the criteria to diagnose the condition of patients suffered from Parkinson’s Disease. The purpose of my research is to design and implement the algorithm on the Leap Motion to detect conditions of all fingers trembling on 3-axis. Hope to provide more precise results for doctors’ diagnosis and evaluation. There are two normal people and two patients as experimental subjects in my research. According to the results of experiments, we can find that not only the frequency of fingers trembling from my system approximates to that from Tremogram, but also my system can quantize the amplitude of fingers trembling on 3-axis. My research successfully presents more diverse information with low cost machine compared with Tremogram. Hope that my system can keep improving to be used on clinical applications in the future.