A Deep Learning Method for Heartbeat Detection in ECG Image

摘要：

　　Although heartbeat segmentation can be done very well in ECG signals for arrhythmia detecting,therere short of techniques for detecting heartbeat part from ECG images.We apply the powerful Faster R-CNN detector here,and achieves accurate detecting results.Along with the improved patch-sampling mechanism in training,detection results are more precise.The high evaluation metric on validation data and demo of real scenes demonstrate the effectiveness of our method.

关键词： Heartbeat detection ECG images Faster R-CNN detector

作者: Zewen He Jinghao Niu Junhong Ren Yajun Shi Wensheng Zhang

作者单位: Institute of Automation,Chinese Academy of Sciences,95 Zhongguancun East Road,Beijing 100190,China;U Institute of Automation,Chinese Academy of Sciences,95 Zhongguancun East Road,Beijing 100190,China Department of Cardiology,Chinese PLA General Hospital,28 Fuxing Road,Beijing 100853,China

会议类型: 国际会议

会议名称: 2019中国智能自动化大会(CIA,2019)

会议地点: 江苏镇江

会议语种:英文

页码: 356-363

在线出版日期: 2019-09-20（万方平台首次上网日期，不代表论文的发表时间）