Stomach tissue classification using autofluorescence spectroscopy and machine learning.

BACKGROUND AND OBJECTIVES: Determination of stomach tumor location and invasion depth requires delineation of gastric histological structure, which has hitherto been widely accomplished by histochemical staining. In recent years, alternative histochemical evaluation methods have been pursued to accelerate intraoperative diagnosis, often by bypassing the time-consuming step of dyeing. Owing to strong endogenous signals from coenzymes, metabolites, and proteins, autofluorescence spectroscopy is a favorable candidate technique to achieve this aim. MATERIALS AND METHODS: We investigated stomach tissue slices and block specimens using a fast fluorescence imaging scanner. To obtain histological information from broad and structureless fluorescence spectra, we analyzed tens of thousands of spectra with multiple machine-learning algorithms and built a tissue classification model trained with dissected gastric tissues. RESULTS: A machine-learning-based spectro-histological model was built based on the autofluorescence spectra measured from stomach tissue samples with delineated and validated histological structures. The scores from a principal components analysis were employed as input features, and prediction accuracy was confirmed to be 92.0%, 90.1%, and 91.4% for mucosa, submucosa, and muscularis propria, respectively. We investigated the tissue samples in both sliced and block forms using a fast fluorescence imaging scanner. CONCLUSION: We successfully demonstrated differentiation of multiple tissue layers of well-defined specimens with the guidance of a histologist. Our spectro-histology classification model is applicable to histological prediction for both tissue blocks and slices, even though only sliced samples were trained.

Spectroscopic Study of Time-Varying Optical Redox Ratio in NADH/FAD Solution.

Autofluorescence imaging has been widely applied as advanced noninvasive diagnostics for in vivo and ex vivo tissues. The optical redox ratio (ORR), which is defined as the fluorescence intensity ratio between reduced nicotine adenine dinucleotide (NADH) and oxidized flavin adenine dinucleotide (FAD), has been used as a diagnostic parameter strongly, because NADH and FAD play an important role in energetic and respiratory metabolism as coenzymes. The ORR method has provided successful assessment in cancer diagnosis including breast, cervical, and oral cancer; few studies have been reported about optical and chemical interference between two molecules resulting in a change in ORR values. In this study, we investigated the variations in ORR values of NADH/FAD mixtures dissolved in tris(hydroxymethyl)aminomethane, phosphate buffer, and deionized water environments. In vitro solutions were prepared in various concentration ratios and the experimental and theoretical ORR values were obtained from fluorescence and absorption spectra in time series. Based on the spectroscopic analysis, we concluded that the inner filter effect causes an instant decrease in FAD fluorescence just after dissolution and that the oxidation-reduction coupled with oxygenation reaction results in time-varying decreases in NADH fluorescence and FAD emission.

Multicolor-Raman analysis of Korean paintworks: emission-like Raman collection efficiency.

It has been reported that the scattering cross-sections of resonance Raman spectra strongly depend on the resonance between the laser's excitation energy and the electronic absorption band of pigments in solution. However, the actual collection of scattered photons is affected by diffuse scattering and self-absorption when studying painted colorants in artworks. Quantitative spectroscopic measurements are required to elucidate the apparent resonance Raman cross-sections in both solution and solid. In this study, we explored the excitation-dependent Raman scattering of natural and artificial Korean pigments painted on a wood block with six visible wavelengths. Our study shows that the Raman intensity profile agrees with the emission profile rather than with the absorption. We also assessed the validity of self-absorption and the outgoing resonance mechanism in the solid state for the results.

Analysis of subcutaneous swine fat via deep Raman spectroscopy using a fiber-optic probe.

Since the fat content of pork is a deciding factor in meat quality grading, the use of a noninvasive subcutaneous probe for real-time in situ monitoring of the fat components is of importance to vendors and other interested parties. In this work, we developed a spectroscopic method using a fiber-optic probe for subcutaneous fat analysis that utilizes spatially offset Raman spectroscopy (SORS). Here, normalized Raman spectra were acquired as a function of spatial offset, and the relative composition of fat-to-skin was determined. We found that the Raman intensity ratio varied disproportionately depending on the fat content and that the variations of the slope were correlated to the thickness of the fat layer. Furthermore, ordinary least square (OLS) regression using two components indicated that the depth-resolved SORS spectra reflected the relative thickness of the fat layer. We concluded that the local distribution of subcutaneous fat could be measured noninvasively using a pair of fiber-optic probes.

Tuning 2D Light Upconversion Emission by Modulating Phonon Relaxation.

The photonic upconversion in rare earth atoms is widely used to convert "invisible" near infrared photons to "visible" photons with continuous wave light. By using a patterned substrate, upconversion become a route for creating new information-incorporating security codes. The amount of information in the cipher increases in proportion to the number of emission colors as well as the pattern structure. Subsequently, changing the chemical composition of upconversion phosphors on 2 D substrates is required to manufacture information-rich upconversion cryptography. In this study, we exploited temperature-controlled thermal reaction on upconversion films deposited on a quartz substrate to prepare security information codes. Multiple color emission was generated from upconversion films as the result of inserting high-frequency molecular oscillators into the film structures. Fourier-transform infrared (FTIR) and time-resolved study corroborated the mechanism of spectral variation of upconversion films.

Spectral evidence for multi-pathway contribution to the upconversion pathway in NaYF4:Yb3+,Er3+ phosphors.

Although upconversion phosphors have been widely used in nanomedicine, laser engineering, bioimaging, and solar cell technology, the upconversion luminescence mechanism of the phosphors has been fiercely debated. A comprehensive understanding of upconversion photophysics has been significantly impeded because the number of photons incorporated in the process in different competitive pathways could not be resolved. Few convincing results to estimate the contribution of each of the two-, three-, and four-photon channels of near-infrared (NIR) energy have been reported in yielding upconverted visible luminescence. In this study, we present the energy upconversion process occurring in NaYF4:Yb3+,Er3+ phosphors as a function of excitation frequency and power density. We investigated the upconversion mechanism of lanthanide phosphors by comparing UV/VIS one-photon excitation spectra and NIR multi-photon spectra. A detailed analysis of minor transitions in one-photon spectra and luminescence decay enables us to assign electronic origins of individual bands in multi-photon upconversion luminescence and provides characteristic transitions representing the corresponding upconversion channel. Furthermore, we estimated the quantitative contribution of multiple channels with respect to irradiation power and excitation energy.