ECG Patterns Recognition using Multilayer Perceptrons and Classification Tree

以多層感知機與分類樹進行心電圖模樣辨識

林堉仁(Yu-Jen Lin) ； 楊景雄(Jing-Xiong Yang)

《電子化優質健康照護》 2007年 (2007/11) Pp. 387-409

https://doi.org/10.29651/ECARE.200711.0387

摘要

本文使用多層感知機(MLP)、分類樹(CT)、以及兩者之結合，進行心電圖四種不同模樣辨識：正常(Normal)、左側東傳導阻滯(LBBB)、右側束傳導阻滯(RBBB)、單純性心室早期收縮(PVC)。我們下載了MIT/BIT的心跳資料庫，加以整理做為我們的訓練與測試資料。我們首先分別使用多層感知機與分類樹，進行這四種心電圖模樣辨識。再來，我們將分類樹執行過程中，用來之分裂節點的輸入屬性揀選起來，做為多層感知機的輸入屬性，再進行模樣辨識。這個做法一來可以降低輸入向量的維度以減少多層感知機計算負擔，二來可以得知心跳訊號有哪些輸入屬性，對於這四種心電圖模樣之辨識較具影響。為了要驗證此方法，我們還使用主成份分析法(PCA)來降低輸入屬性維度，再進行多層感知機之心電圖模樣辨識。經過多次的電腦程式模擬與計算，辨識的結果證實了我們提出方法的可行性與優越性。

關鍵字

心電圖 ； 模樣辨識 ； 多層感知機 ； 分類樹 ； 主成份分析

並列摘要

This paper presents an approach based on the combination of multilayer perceptrons (MLP) and classification tree (CT) to recognising four electrocardiograms (ECG) patterns: normal, left bundle branch block (LBBB), right bundle branch block (LBBB) and premature ventricular contraction (PVC). This study utilises MIT/BIH arrhythmia database as training and testing data. We first apply MLP and CT to recognise ECG patterns. Next, we collect the ECG signal features that are selected in splitting CT's terminal node, and feed these selected features into MLP for ECG pattern recognition. The aim is twofold: reducing the input attributes of MLP so as to lower computation burden, and understanding which heartbeat features are crucial in above four ECG patterns recognition. In order to compare the effectiveness of proposed method, the principal component analysis (PCA) technique is also applied to cut down the input feature dimension for ECG pattern recognition. Comprehensive computer simulations in the end of this paper will justify the feasibility and superiority of the proposed approach.

並列關鍵字

Electrocardiograms ECG ； Pattern Recognition ； Multilayer Perceptrons MLP ； Classification Tree CT ； Principal Component Analysis PCA