Most biomedical applications of terahertz (THz) imaging are based on distinguishing the dielectric differences of tissues or cells in the THz band. But changes in bio-sample dehydration during the point-scanning process can lead to time-varying deviations in the imaging results. To eliminate the deviations, we proposed a time-varying dehydration kinetic model by analyzing the water diffusion process. The model is verified by experiments and applied to restore each point close to the initial imaging starting state of fresh cellular samples, compensating for the impact of slow speed point-scanning on image deterioration. This methodology has significantly improved the THz super-resolution imaging quality of fresh cellular samples, and successfully restored the cell contours that had been obscured by the cell dehydration over time. Although the dehydration model is developed in THz near-filed imaging, it also pertains to other spectral systems that operate in the raster-scan mode on fresh bio-samples.
Keywords: Cell dehydration model; Correction algorithm; Living biological samples; THz near-filed scanning imaging.
Copyright © 2023 Elsevier B.V. All rights reserved.