Spectral Biopsy of Brain Function via Wavelet Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for determining intrinsic neuronal activity, such as Hilbert transform-based phase-amplitude coupling, are limited by assumptions that natural neuronal signals often violate, making them unsuitable for diagnosing neurological and psychiatric diseases.
Innovation Solution
A computer-implemented method that extracts wideband low-frequency signals and broadband gamma envelope signals from brain activity measurements, calculating cross-correlation to produce Tau Modulation Curves, which indicate the modulation of broadband gamma activity without requiring signals to be within narrow frequency bands or sinusoidal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If Hilbert transform-based phase-amplitude coupling is used to analyze neuronal activity, then the method provides a structured approach to extract amplitude and phase components, but it requires filtering signals into narrow frequency bands which violates the natural broadband nature of neuronal signals
Solution Approach 1:
The patent changes the fundamental parameters of the analysis method by replacing the Hilbert transform with a wavelet-based approach. This allows the method to operate on broadband signals without requiring narrowband filtering, thereby maintaining adaptability to natural neuronal signals while providing structured analysis through wavelet coefficients and phase-amplitude coupling metrics.
Solution Approach 2:
The patent substitutes the mathematical mechanism of the Hilbert transform with a wavelet transform-based mechanism. This replacement enables the extraction of amplitude and phase information from broadband signals without the need for frequency band filtering, resolving the contradiction between structured analysis and broadband signal compatibility.
2Ease of manufacture
If traditional phase-amplitude coupling methods are applied, then the analysis follows a standardized three-step process, but it assumes neuronal activity is sinusoidal and sustained which natural signals often violate
Solution Approach 1:
The patent changes the mathematical parameters from Hilbert-based sinusoidal assumptions to wavelet-based time-frequency representations. This allows the method to accommodate non-sinusoidal, transient, and aperiodic characteristics of natural neuronal signals while maintaining a standardized analysis workflow through wavelet decomposition and coupling metric calculation.
Solution Approach 2:
The patent introduces dynamic adaptability by using wavelet transforms that can capture transient and non-stationary features of neuronal signals. The method dynamically adjusts to signal characteristics through multi-resolution analysis, allowing reliable detection of phase-amplitude coupling in natural signals that violate traditional sinusoidal and sustained assumptions.
3Measurement precision
If narrowband filtering is applied to extract phase information, then the Hilbert transform can be applied, but it loses broadband frequency information and creates arbitrary band definitions
Solution Approach 1:
The patent segments the frequency analysis into multiple wavelet scales rather than arbitrary narrow bands. This segmentation preserves broadband information by analyzing multiple frequency ranges simultaneously through wavelet decomposition, allowing phase extraction at different scales without losing overall broadband spectral content.
Solution Approach 2:
The patent transitions from one-dimensional frequency filtering to a two-dimensional time-frequency representation using wavelet transforms. This dimensional change allows simultaneous access to both temporal and spectral information, enabling phase extraction while preserving broadband frequency content that would be lost in traditional narrowband filtering approaches.
Data Source
AI summary
Methods and systems are disclosed for analyzing interactions between low-frequency oscillations and high-frequency activity in electromagnetic brain signals such as EEG, MEG, SEEG, and ECoG signals in subjects in real-time that does not depend on the signals being contained within narrow frequency bands, sinusoidal, sustained and monolithic. The disclosed methods and systems can be applied to electromagnetic brain signals to detect brain activity alterations associated with neurological and psychiatric diseases.