Use of a priori information in estimating tissue resistivities--a simulation study

Area of Science:

• Bioengineering
• Medical Imaging
• Computational Electrophysiology

Background:

• Accurate in vivo tissue resistivity estimation is crucial for developing reliable human body volume conductor models.
• Electrical impedance tomography (EIT) techniques, specifically four-electrode impedance measurements, provide data for resistivity estimation.

Purpose of the Study:

• To introduce a novel resistivity estimation algorithm, the statistically constrained minimum mean squares error estimator (MiMSEE), to enhance accuracy.
• To incorporate a priori geometrical information, statistical properties of regional resistivities, linearization error, and instrumentation noise into the estimation algorithm.

Main Methods:

• MiMSEE utilizes geometrical information from high-resolution imaging modalities.
• Simulated measurements were generated from numerical models with five and six regions.
• The performance was evaluated by increasing the number of current electrode pairs and compared against a conventional least squares error estimator (LSEE).

Main Results:

• MiMSEE achieved up to 27 times smaller estimation error compared to LSEE, particularly for small, high-contrast regions like the aorta.
• The MiMSEE algorithm requires more computational time (25.8x for 5 regions, 22.2x for 6 regions) than LSEE.
• The computational time difference decreases as the number of regions increases.

Conclusions:

• The MiMSEE algorithm offers a significant improvement in accuracy for in vivo tissue resistivity estimation.
• The trade-off between accuracy and computational cost should be considered for practical applications.
• Further research can explore optimizing MiMSEE for complex multi-region models.