Parallel ECG Signal Analysis Using Integrated CPU and GPU
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional cloud platform-based ECG signal analysis faces challenges in real-time response due to high computational demands, and mobile terminals with limited CPU performance struggle to handle long-term ECG signal processing efficiently, leading to delayed feedback and increased power consumption.
Innovation Solution
A parallel analysis apparatus and method utilizing a mobile terminal with integrated CPU and GPU, where the CPU performs primary noise reduction and initial classification, and the GPU performs characteristic extraction and secondary noise reduction, enabling template matching classification to improve analysis efficiency and timeliness while reducing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If ECG signal analysis is performed on a mobile terminal using traditional CPU-based sequential processing, then the device can handle long-term ECG signals, but the analysis speed is slow and real-time response is difficult to achieve
Solution Approach 1:
The patent segments the ECG analysis workload by dividing it into two parts: preliminary screening performed by the CPU and detailed analysis performed by the GPU. This segmentation allows the CPU to handle basic processing tasks while the GPU accelerates computationally intensive operations, thereby improving overall analysis speed without overwhelming the mobile terminal's processing capabilities
Solution Approach 2:
The patent introduces an integrated memory as an intermediary between the CPU and GPU, enabling efficient data sharing and communication. This intermediary structure allows both processors to access the same data without conflicts, facilitating smooth collaboration and improving processing capability while maintaining fast access to ECG signal data
2Productivity
If ECG signal analysis is performed on a mobile terminal with limited CPU performance, then the device can operate portably, but the processing capability is insufficient for long-term ECG signals
Solution Approach 1:
The patent makes the mobile terminal's hardware configuration universal by integrating both CPU and GPU processors that can handle different types of computational tasks. The CPU handles control and preliminary processing, while the GPU handles intensive analysis, creating a multi-functional processing system that enhances overall processing capability without requiring separate dedicated devices
Solution Approach 2:
The patent applies preliminary action by having the CPU perform preliminary screening and noise reduction on ECG signals before passing the preprocessed data to the GPU for detailed analysis. This preliminary processing reduces the computational burden on the GPU and improves overall processing capability by eliminating unnecessary data early in the analysis pipeline
3Reliability
If traditional cloud platform-based ECG signal analysis is used, then computing power is sufficient, but real-time response is difficult in unstable network environments
Solution Approach 1:
The patent enables the mobile terminal to perform ECG signal analysis independently using its integrated CPU and GPU, without requiring continuous connection to external cloud platforms. This self-service capability ensures reliable real-time response even in unstable network environments, as the device can process and analyze ECG data locally without external assistance
Solution Approach 2:
The patent transitions from cloud-based remote processing to local on-device processing by adding GPU computing dimension to the mobile terminal. This dimensional change in processing architecture enables the device to handle computationally intensive ECG analysis locally, improving response timeliness and reliability while reducing dependence on network connectivity
Data Source
Figure 1~2
Figure 3~4
AI summary
Present disclosure provides an ECG signal parallel analysis apparatus, method and a mobile terminal. The apparatus includes: an integrated memory, a CPU and a GPU. The integrated memory includes a first memory for being used by the CPU and a second memory for being used by the GPU, and the CPU may access the second memory. The CPU performs primary noise reduction on a received ECG original signal to obtain a primary ECG signal, and performs abnormal heartbeat classification preliminary screening on characteristic data extracted from the GPU to obtain suspected abnormal heartbeat data. The GPU performs characteristic extraction on the primary ECG signal to obtain characteristic data, performs secondary noise reduction on the primary ECG signal to obtain a secondary ECG signal, and processes the suspected abnormal heartbeat data and the secondary ECG signal by applying a template matching classification mode to obtain final abnormal heartbeat data. Present disclosure may increase efficiency of ECG signal analysis, and thus improves the timeliness of ECG signal analysis and feedback.