@article{oai:hama-med.repo.nii.ac.jp:00000524,
 author = {澤田, 成生 and 小山, 孝夫 and 木村, 元彦 and 水品, 静夫 and 木村, 泰三 and 原田, 幸雄 and 杉浦, 敏文},
 issue = {404},
 journal = {電子情報通信学会技術研究報告. MBE, MEとバイオサイバネティックス IEICE technical report. ME and bio cybernetics},
 month = {Dec},
 note = {本報告では,致死性不整脈である心室細動と正常心拍の判別を心電図の特徴に着目して行う方法について述べる. 心電図の特徴を表す量として.(1)心周期の変動率,(2)振幅変動率,(3)波形のピークの尖度,の三つに着目,各量を定義し,標準的な心電図を用いてそれらをを求めた.これら用いてファジイ推論による心電図評価法を開発し,正常から心室細動までの範囲を100〜0の数値に対応させた結果,正常な心電図はほぼ90以上に,細動波形は20以下の数値に対応した.心電図波形を簡単な認識アルゴリズムによって見分ける可能性か示唆された.  A study of the method to discriminate the ventricular fibrillation from the normal sinus rhythm based on the different of the electrocardiogram (ECG) characteristics is presented. Three characteristics, (1) fluctuation rate of the interval, (2) fluctuation rate of the amplitude, (3) sharpness of the peak, are defined. These were measured with three normal ECGs and three ECGs of the ventricular fibrillation. Based on this result a method to evaluate ECGs through fuzzy inference was developed. ECGs are quantified by the numbers, 0(fibrillation)〜100(normal). Seven ECGs were measured by this method. ECGs over 90 were determined to be normal while those less than 20 were to be fibrillation. Feasibility for the discrimination of cardiac arrhythmia by this simple method is discussed.},
 pages = {75--80},
 title = {不整脈の自動検知に関する研究},
 volume = {95},
 year = {1995}
}