A comparison of two-dimensional techniques for converting magnetocardiogram maps into effective current source...

Area of Science:

• Biophysics
• Biomedical Engineering
• Medical Imaging

Background:

• Current-arrow maps (CAM) provide pseudo-two-dimensional current data.
• Accurate physical current distribution is crucial for interpreting physiological signals like magnetocardiography (MCG).
• Existing methods require conversion of CAM data to physical current for detailed analysis.

Purpose of the Study:

• To develop and validate a method for converting pseudo-2D CAM current data into physical current distributions.
• To establish a relationship between CAM data and physical current using Tikhonov regularization.
• To determine the optimal regularization parameter for accurate current reconstruction in MCG analysis.

Main Methods:

• Optimal solution in a least mean square sense with Tikhonov regularization (LMSTR) was employed.
• Current dipole simulations were used to calculate current pattern differences (ΔJ) with varying regularization parameters (α).
• Magnetocardiographic (MCG) signals from healthy subjects were analyzed, correlating CAM and LMSTR current values, and regression analysis was performed against reconstruction depth (z(d)).

Main Results:

• Minimal current pattern differences (ΔJ) were observed with a regularization parameter α ≈ 10⁻¹⁰.
• This optimal parameter (α = 10⁻¹⁰) yielded minimal ΔJ at key points (P-wave, QRS-complex, T-wave peaks) in MCG signals.
• High correlation coefficients (>0.9) were found between CAM and LMSTR current values, with regression slopes (y) strongly correlated with reconstruction depth (z(d)) (r = -0.93).

Conclusions:

• The study successfully established a conversion method from CAM pseudo-2D current to LMSTR physical current.
• The conversion factor is the slope (y) derived from the regression line, which is dependent on the reconstruction depth (z(d)).
• This depth-dependent conversion allows for the transformation of CAM results into effective physical current distributions for improved MCG analysis.