Investigation of the effects of moisture content using mid-infrared passive spectroscopic imaging. [Not invited]
矢野響; 石丸伊知郎
European Japan Experts Association 2025 2025/09 信州大学 European Japan Experts Association
We are developing medical applications for mid-infrared passive spectroscopic imaging, a technique that directly analyzes radiation centered around 10 μm emitted from the human body (~300 K). This approach enables non-invasive, stand-off acquisition of intrinsic spectral characteristics. In contrast, conventional Fourier Transform Infrared Spectroscopy (FTIR) is an active method that illuminates the sample with infrared light and measures the transmitted or reflected signal. While widely used, FTIR is limited in applicability to moisture-rich samples such as biological tissues and typically probes only several micrometers in depth. Mid-infrared passive spectroscopic imaging, by directly measuring the target’s emission, integrates spectral information through the entire optical path, allowing access to deeper-layer composition. Using this method, we have detected glucose and lactic acid in vivo. However, the effect of sample moisture content on passive spectroscopy performance remains unclear. In this study, we systematically investigate and evaluate the influence of moisture content on mid-infrared passive spectroscopic imaging. First, aiming to apply the proposed method to medical use, we measured the liver using both FTIR and passive spectroscopy. The dotted line in Figure 1(a) shows the spectral characteristics of the liver measured by FTIR, where a characteristic absorption peak at 9.7 µm was observed. However, due to infrared absorption of water, spectral characteristics above 11 µm could not be confirmed. In contrast, the solid line in Figure 1(a) represents the spectral characteristics obtained using passive spectroscopy. In this experiment, the liver was heated for 0 to 25 minutes to dry it. As a result, an emission peak at 9.7 µm was observed throughout the 0–25 minute period, like FTIR. Moreover, spectral characteristics beyond 11 µm, which could not be measured by FTIR, were successfully obtained. Next, the measurement depth of passive spectroscopy was examined using a liver sample cut diagonally, as shown in Figure 1(b). The thinnest part of the sample was approximately 1 mm thick, and the thickest about 4 mm. The measurement results are shown at the bottom of Figure 1(b). At the thinnest part, a peak at 9.7 µm was observed, but at the other four points, the spectral characteristics deteriorated. These results indicate that qualitative information on components at a depth of about 1 mm can be obtained. This suggests that passive spectroscopy has the potential to detect components at depths of about 1 mm in biological measurements. We have already confirmed that applying a temperature gradient to the measurement target facilitates acquisition of spectroscopic characteristics. Building on this understanding of moisture content, sample thickness, and temperature gradients, we aim to achieve quantitative medical measures using mid-infrared passive spectroscopic imaging.