Dynamic Lactate Monitoring Method Based on Spatiotemporal Inward Bioimpedance Spectroscopy Tomography

This study introduces a novel dynamic lactate monitoring approach using spatiotemporal-inward bioimpedance spectroscopy tomography (SI-BIST) to assess lactate distribution changes in human tissue. Numerical simulations were conducted to evaluate the influence of different electrode configurations and measurement patterns on the accuracy of lactate monitoring. Three electrode configurations were proposed: ${R}_{\text {16-16}}$ (two rings with 16 electrodes each), ${R}_{\text {8-16-8}}$ (three rings with 8, 16, and 8 electrodes, respectively), and ${R}_{\text {8-8-8-8}}$ (four rings with eight electrodes each). Additionally, four measurement patterns, namely, adjacent, opposite, “zigzag,” and “snake-shaped” methods, were analyzed to determine the optimal measurement pattern. The imaging performance of the SI-BIST method was further investigated for lactate targets of varying sizes under different interlayer spacings between electrode rings. The numerical simulation results showed that the ${R}_{\text {16-16}}$ electrode configuration combined with the “snake-shaped” measurement pattern provided the superior imaging performance achieving an average image correlation coefficient (ICC) of 0.84. Under this optimal setup, the lactate volume ( ${V}_{\text {stage}}$ ) and barycenter offset ( ${D}_{\text {off}}$ ) demonstrated accurate and reliable measurements. Bioimpedance spectroscopy (BIS) simulation results indicated that relaxation impedance decreased with increasing lactate concentration, validating the capability of BIS method to differentiate lactate concentrations. In the experimental studies, the SI-BIST platform successfully monitored lactate diffusion and classified concentrations. Reconstructed images showed that lactate volume ( ${V}_{\text {stage}}$ ) in reconstructed images increased from 1093 in Stage 1 to 4843 in Stage 5, while the spatial-mean conductivity ( $\sigma _{\text {stage}}$ ) rose from 0.022 to 0.061 S/m as lactate diffusion time (T) progressed. Measurements using agar models with varying lactate concentrations achieved a classification accuracy of 74% with a weighted k-nearest neighbor (KNN) algorithm model. This study establishes SI-BIST as an effective technique for monitoring spatiotemporal variations in lactate concentration, offering robust capabilities for imaging lactate diffusion and accurately classifying concentrations.