Adaptive Time-Frequency Windows for iEEG Seizure Onset Zone Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying the seizure onset zone (SOZ) in epilepsy are challenging due to the complex localization of the SOZ, variable seizure patterns, and the need for invasive recording techniques. Existing biomarkers often rely on predefined frequency bands and do not adapt to patient-specific patterns, lacking a fully unsupervised algorithm for SOZ localization.
Innovation Solution
A computer-implemented method that identifies time-frequency features of physiological events by filtering and analyzing intracranial electroencephalography (iEEG) signals within defined time-frequency windows. This method calculates feature values and quantifiers, such as global activation and activation entropy, to select relevant time-frequency windows and identify the seizure focus.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If predefined frequency bands are used for biomarker quantification, then the analysis framework is simplified, but patient-specific seizure patterns cannot be adequately captured
Solution Approach 1:
The patent transitions from static predefined frequency bands to dynamic adaptive time-frequency windows. The system automatically adjusts the time-frequency regions of interest based on detected seizure patterns, allowing the analysis framework to adapt to patient-specific characteristics while maintaining computational feasibility through algorithmic automation.
Solution Approach 2:
The invention changes the parameters of frequency analysis by moving from fixed frequency bands to variable time-frequency windows. The system identifies and analyzes specific time-frequency regions where seizure-related power changes occur, adapting the frequency and temporal parameters to match actual seizure patterns rather than forcing data into predetermined categories.
2Extent of automation
If fully unsupervised algorithms are implemented for SOZ localization, then automation is improved, but reliability of detection decreases due to lack of expert validation
Solution Approach 1:
The system incorporates feedback mechanisms where detection results inform subsequent analysis. The automated algorithm uses detected power changes and time-frequency patterns to refine its own detection criteria, and the system provides feedback to clinicians through visualizations that can be validated against expert knowledge, creating a loop that improves both automation and reliability.
Solution Approach 2:
The patent implements self-service through automated detection algorithms that independently identify seizure patterns and localize SOZ without requiring continuous expert intervention. The system serves itself by automatically adjusting analysis parameters, selecting time-frequency windows, and generating localization results, reducing dependency on manual expert validation while maintaining reliability through multiple detection criteria.
3Measurement precision
If multiple time-frequency windows are analyzed, then detection sensitivity is improved, but computational complexity increases
Solution Approach 1:
The patent segments the frequency spectrum into multiple time-frequency windows that can be analyzed independently. Each window focuses on specific frequency ranges and time periods where seizure-related activity is most likely to occur, allowing detailed analysis without requiring exhaustive processing of the entire signal spectrum simultaneously.
Solution Approach 2:
The system applies partial action by concentrating computational resources on only those time-frequency windows that show power changes consistent with seizure activity. Rather than uniformly analyzing all possible time-frequency combinations, the algorithm identifies and focuses on the most relevant windows, reducing overall computational complexity while maintaining high detection sensitivity.
Data Source
Figure 1
Figure 2A~2B
Figure 2C(A)~2C(D)
AI summary
A method and computer programs for identifying time-frequency features of physiological events are disclosed. A computer system comprises filtering a set of physiological signals within each one of a plurality of time-frequency windows, obtaining a filtered set for each time-frequency window; calculating, for each time- frequency window, a given feature for the filtered set, each one of the signals of the filtered set having a given feature value, providing for each time-frequency window a set of feature values; and calculating, for each time-frequency window a first quantifier defined as a function of said set of features values and/or a second quantifier defined as a function of an empirical distribution of said set of feature values. The first quantifier can be compared with a first threshold and the second quantifier can be compared with a second threshold. The computing system can further select the time-frequency windows satisfying the first threshold and/or the second threshold.