DenseNet Dual-Channel LSTM for QRS Complex Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current deep learning methods for detecting QRS complexes in electrocardiogram signals face challenges due to complex signal shapes and individual differences in waveforms, resulting in low detection performance.
Innovation Solution
A method utilizing a detection model comprising a DenseNet and a dual-channel Long Short-Term Memory (LSTM) is proposed, where time sequence data is input to extract spatial features, and these features are fused using an attention mechanism within the dual-channel LSTM to improve detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If deep learning method is used for QRS complex detection, then detection speed is improved, but detection accuracy deteriorates due to complex signal shapes and individual differences
Solution Approach 1:
The patent segments the detection task into two independent channels: a first channel for processing original time sequence data and a second channel for processing extracted spatial features. This segmentation allows each channel to specialize in specific aspects of QRS detection, with the first channel capturing temporal patterns and the second channel capturing spatial characteristics, thereby improving overall detection accuracy while maintaining speed
Solution Approach 2:
The patent introduces a spatial feature extraction dimension by applying a one-dimensional convolutional neural network to the time sequence data. This transforms the one-dimensional time series into two-dimensional spatial feature maps, enabling the dual-channel LSTM to process both temporal and spatial information simultaneously, which resolves the accuracy limitation of traditional single-channel deep learning methods
2Device complexity
If single-channel LSTM is used, then model complexity is reduced, but feature extraction capability deteriorates
Solution Approach 1:
The patent merges two independent LSTM channels into a unified detection system where the first channel processes original time sequence data and the second channel processes spatial features. The outputs of both channels are then fused through attention mechanism and concatenation, combining the advantages of both simple temporal processing and complex spatial feature extraction to achieve superior detection performance
Solution Approach 2:
The patent introduces an attention mechanism as an intermediary component that weights and combines the outputs from the first and second LSTM channels. This intermediary allows the model to dynamically adjust the importance of different feature channels based on the input signal characteristics, enhancing feature extraction capability while maintaining manageable model complexity through the modular architecture
Data Source
AI summary
A method for detecting a QRS complex of an electrocardiogram signal, an apparatus therefor, a device and a medium are provided. A QRS complex of an electrocardiogram signal is detected by a detection model including a DenseNet and a dual-channel Long Short-Term Memory (LSTM). Spatial feature information of the QRS complex in the electrocardiogram signal is learned through the DenseNet, and then the spatial feature learned by the DenseNet and the time sequence data are input into respective channels of the dual-channel LSTM, so that the dual-channel LSTM can fuse the spatial information and the time sequence information of the QRS complex in the electrocardiogram signal, thereby improving a segmentation effect of the model on features, and finally improving accuracy of detecting the QRS complex.