心血管疾病目前是世界上最常見的死因之一。心電圖訊號是判斷心臟疾病的重要依據,並即時為病人提供醫療資源。但是這些心電圖資訊很可能因為器材操作上的錯誤、電極的感應不良、測量環境的不適當、甚至是病人的呼吸而產生雜訊。這些雜訊可能令醫生產生錯誤的診斷,對許多心電圖自動判讀系統而言,這些雜訊會讓系統擷取出錯誤的心電圖資訊,也會影響判斷疾病的正確率。所以偵測與去除心電圖中的雜訊變成一項重要的題目。在本篇研究中,我們提出了六種常見於心電圖中的雜訊型態,分別是平坦、陡峭、飽和、低震幅、相似正弦波、基線飄移等六種,每一種都分別經由不同的訊號前處理來偵測,並且針對無法復原或無法得到心電圖資訊的雜訊片段進行消除與重新組合,讓有雜訊的資料的易讀性更高。
The Cardiovascular disease (CVDs) is one of the most common cause of death in the world. The analysis of Electrocardiograms (ECGs) is a important tools in early diagnosis arrhythmias. However sometime these measurement data would be corrupted by noises which may cause by the wrong equipment operation, poor contact of the electrode, or even the breath of the patients. These noises would make cardiologists or automatic CVDs detection system hard to make a correct diagnosis. Therefor, the noise detection and elimination from ECG data become an important project on Health Information System (HIS). In this study, we propose six most common types of noise. For each noise type detection, we apply difference signal preprocessing. If there exist some noise segments that have no information and can not be repaired, we will eliminate them and combine the remain usable segments into a complete signal.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。