Record Linkage Using Temporal Power Spectrum Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing record linkage methods, such as the Fellegi-Sunter approach, fail to accurately match patient records due to issues like false-negatives and false-positives, excessive reliance on variables, and inconsistencies in patient identifiers, leading to incomplete data sharing and potential errors in healthcare.
Innovation Solution
Incorporating power-spectrum-based temporal pattern-specific weighting into record linkage methods, using Bayesian power spectra analysis to enhance accuracy by calculating value-specific weights and leveraging time-series data for improved matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional record linkage methods (Fellegi-Sunter) are used, then the matching process is simple and fast, but the accuracy is insufficient due to false-negatives and false-positives
Solution Approach 1:
The patent transforms temporal pattern information from the time domain to the frequency domain using power spectrum analysis. This parameter transformation enables the system to capture temporal dynamics that are invisible in the time domain, thereby improving match discrimination without proportionally increasing system complexity
Solution Approach 2:
The patent adds a frequency domain dimension to the traditional time-domain record linkage approach. By computing power spectra of temporal patterns and comparing frequency-domain characteristics, the system creates a new dimension of discrimination that reduces false matches while maintaining computational feasibility
2Measurement precision
If more variables are used in record linkage, then the matching accuracy may improve, but the computational complexity and data requirements increase excessively
Solution Approach 1:
The patent extracts and isolates the temporal pattern component from patient records, treating it as a separate feature dimension. By computing power spectra specifically for temporal intervals between episodes, the system extracts discriminative information without requiring additional clinical variables, thus improving accuracy without proportionally increasing complexity
3Loss of information
If traditional field-specific weighting is used, then the matching process is straightforward, but value-specific information and temporal patterns are not leveraged
Solution Approach 1:
The patent performs preliminary power spectrum analysis on temporal patterns before the actual record linkage comparison. By pre-computing frequency-domain representations of temporal intervals and storing them as features, the system prepares discriminative information in advance, enabling more informed matching decisions without increasing the complexity of the core linkage algorithm
Data Source
AI summary
Methods, systems, and computer-readable media are provided for facilitating record matching and entity resolution and for enabling improvements in record linkage. A power-spectrum-based temporal pattern-specific weight may be incorporated into record linkage methods to enhance the record linkage accuracy and statistical performance. For example, in embodiments, a value-specific weight may be calculated from a population-based frequency of field-specific values and provides an opportunity to capture and measure the relative importance of specific values found in a field. A timeseries-derived Bayesian power spectrum weight may be calculated from the population-based frequency of temporal pattern-specific values in terms of intensities at various frequencies of the power spectrum computed from the timeseries, and further provides an opportunity to capture and measure the relative importance of specific sequences of care episodes.