發(fā)布時(shí)間：2018-06-01 05:10

  本文選題：深度信念網(wǎng) + 心電節(jié)拍分類��； 參考：《計(jì)算機(jī)工程與設(shè)計(jì)》2017年05期

【摘要】：提出一種基于深度信念網(wǎng)(deep belief network,DBN)和心電波形采樣的心電自動(dòng)分類算法。對(duì)心電信號(hào)進(jìn)行濾波、R波定位后,以QRS波群的180Hz下采樣表示心拍形態(tài),結(jié)合RR間期特征,使用的DBN共6層,隱藏層神經(jīng)元數(shù)目為30。使用標(biāo)準(zhǔn)數(shù)據(jù)庫(kù)對(duì)DBN進(jìn)行訓(xùn)練和測(cè)試,結(jié)果為平均Se88.6%,平均P~+62.1%,優(yōu)于現(xiàn)有特征選擇方法的結(jié)果,基于深度學(xué)習(xí)的心拍分類算法無(wú)需波形特征提取步驟,解決了目前的波形特征對(duì)心拍的人間差異沒(méi)有魯棒性的問(wèn)題。
[Abstract]:An automatic ECG classification algorithm based on deep belief network (DBN) and ECG waveform sampling is proposed. After filtered R wave location of ECG signal, the beat morphology was represented by 180Hz downsampling of QRS wave group. Combined with RR interval characteristics, six layers of DBN were used, and the number of neurons in hidden layer was 30. The standard database is used to train and test the DBN. The results show that the average Se88.6 and P62.1 are better than the existing feature selection methods. The beat classification algorithm based on depth learning does not need waveform feature extraction step. It solves the problem that the current waveform features are not robust to the human differences.
【作者單位】： 中國(guó)科學(xué)院微電子研究所;中國(guó)科學(xué)院微電子研究所昆山分所;
【基金】：國(guó)家自然科學(xué)基金項(xiàng)目(61271423)
【分類號(hào)】：R540.4;TN911.7
，

本文編號(hào)：1962958