ECG Signal Epoch Scoring for Continuous Quality Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Assessing the quality of extensive electrocardiogram (ECG) data is a laborious task, especially in clinical trials, where robust and high-quality ECG recordings are crucial for reliable drug development and cardiac monitoring, but current methods lack efficient automation and precision.
Innovation Solution
A system and method using machine learning (ML) and deep learning (DL) neural networks to decompose ECG signals into epochs, generate quality scores, and produce visualizations and alerts to identify and correct inconsistencies in ECG data, utilizing convolutional neural networks (CNNs) and rolling windows to compute local and occurrence metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual review methods are used to assess ECG data quality, then measurement precision can be maintained, but productivity and time efficiency deteriorate significantly
Solution Approach 1:
The patent replaces manual mechanical review processes with an automated machine learning-based system. The ML model automatically processes ECG signals, decomposes them into epochs, generates quality scores, and identifies artifacts, eliminating the need for manual inspection while maintaining assessment accuracy.
Solution Approach 2:
The system performs self-service by automatically assessing its own output quality through the ML model's quality scoring mechanism. The system self-evaluates ECG data quality without requiring external manual review, generating quality scores and alerts autonomously.
2Quantity of substance
If extensive continuous ECG data is collected for later analysis, then the quantity of available data increases, but the time required for quality checking and data preparation increases proportionally
Solution Approach 1:
The patent applies preliminary action by performing quality assessment and artifact identification during the data collection process itself, rather than waiting for later analysis. The system continuously monitors ECG data, decomposes it into epochs, generates quality scores in real-time, and identifies artifacts as they occur, enabling proactive data management.
Solution Approach 2:
The system maintains continuous operation by processing ECG data streams continuously rather than in batches. The rolling window approach allows continuous decomposition and quality scoring of ECG segments, ensuring that quality assessment is an ongoing process that accompanies data collection without interruption.
3Productivity
If automated quality assessment systems are implemented, then productivity improves, but device complexity and system requirements increase
Solution Approach 1:
The patent applies segmentation by dividing the continuous ECG signal into discrete epochs of fixed length. This segmentation allows the complex quality assessment task to be broken down into manageable units that can be processed independently through the ML model, simplifying the overall system architecture while maintaining high processing speed.
Data Source
AI summary
Exemplary system and methods for assessing ECG data quality. The system includes a processor that decomposes a continuous ECG signal into multiple epochs and generates a quality score for each epoch. A first rolling window of a first specified width is applied to the decomposed ECG signal to capture plural quality scores from a sequence of epochs in the plural epochs. The processor computes a local metric from the plural quality scores captured by the first rolling window. The processor captures plural quality scores and computes associated local metrics across an entirety of the decomposed ECG signal. Once completed the processor executes one or more application modules for: generating a quality alert signal and generating a quality visualization signal for displaying a color map corresponding to signals associated with a local metric. The output of the one or more executed application modules is passed to a user interface.