Source imaging method based on diagonal covariance bases and its applications to OPM-MEG

Area of Science:

• Neuroscience
• Biophysics
• Medical Imaging

Background:

• Magnetoencephalography (MEG) is a noninvasive brain imaging technique.
• MEG source estimation faces challenges due to the ill-posed nature of inverse problems.
• Accurate estimation of source extent is vital for understanding brain function and disease.

Purpose of the Study:

• To develop an improved method for estimating the spatial extent of MEG sources.
• To enhance the accuracy of source localization and extent determination in MEG data.
• To introduce a novel Bayesian framework incorporating spatial constraints for source imaging.

Main Methods:

• Developed the Source Imaging with Diagonal Covariance Bases (SI-DCB) method.
• Modeled the spatial properties of sources using diagonal covariance bases (DCB).
• Employed convex analysis for parameter estimation within a Bayesian framework.

Main Results:

• SI-DCB significantly outperformed five benchmark methods in simulations for estimating patch source location and extent.
• Validated the effectiveness of SI-DCB using somatosensory stimulation data from an OPM-MEG system.
• Demonstrated accurate identification of brain response source areas by SI-DCB.

Conclusions:

• SI-DCB offers a superior approach for accurate MEG source extent estimation.
• The method provides a template for enhancing source imaging accuracy within sparse Bayesian frameworks.
• SI-DCB has potential applications in neuroscience research and clinical diagnostics.