Pengyao Xu

Title: Innovations of Pengyao Xu in Electrocardiogram Signal Classification

Introduction

Pengyao Xu is a notable inventor based in Jinan, China. He has made significant contributions to the field of electrocardiogram (ECG) signal classification, holding a total of 8 patents. His work focuses on improving the accuracy and stability of ECG signal analysis through innovative methods.

Latest Patents

One of his latest patents is a few-shot electrocardiogram (ECG) signal classification method based on an improved Siamese network. This method constructs a CMP module as a sub-network of a Siamese network, combining extracted local and global features to analyze peak information such as position, amplitude, and offset. This approach results in a more robust transformed feature vector, enhancing the accuracy and stability of few-shot ECG signal classification.

Another significant patent is an electrocardiogram (ECG) signal classification method based on contrastive learning and multi-scale feature extraction. This method utilizes a SE-ResNeXt-CAN network model, which includes various modules for feature extraction and optimization. By adaptively learning correlations between channels and expanding the receptive field, this model improves the performance and generalization ability of ECG signal classification tasks.

Career Highlights

Pengyao Xu has worked at prominent institutions such as Qilu University of Technology and the Shandong Computer Science Center, which is the National Supercomputing Center in Jinan. His experience in these organizations has contributed to his expertise in the field of ECG signal classification.

Collaborations

He has collaborated with notable colleagues, including Minglei Shu and Zhaoyang Liu, further enhancing his research and development efforts in the field.

Conclusion

Pengyao Xu's innovative approaches to ECG signal classification demonstrate his commitment to advancing medical technology. His contributions are paving the way for more accurate and reliable ECG analysis methods.