Real-time molecular imaging of near-surface tissue using Raman spectroscopy

From Genotype to Phenotype: Raman Spectroscopy and Machine Learning for Label-Free Single-Cell Analysis

Raman spectroscopy has made significant progress in biosensing and clinical research. Here, we describe how surface-enhanced Raman spectroscopy (SERS) assisted with machine learning (ML) can expand its capabilities to enable interpretable insights into the transcriptome, proteome, and metabolome at the single-cell level. We first review how advances in nanophotonics-including plasmonics, metamaterials, and metasurfaces-enhance Raman scattering for rapid, strong label-free spectroscopy. We then discuss ML approaches for precise and interpretable spectral analysis, including neural networks, perturbation and gradient algorithms, and transfer learning. We provide illustrative examples of single-cell Raman phenotyping using nanophotonics and ML, including bacterial antibiotic susceptibility predictions, stem cell expression profiles, cancer diagnostics, and immunotherapy efficacy and toxicity predictions. Lastly, we discuss exciting prospects for the future of single-cell Raman spectroscopy, including Raman instrumentation, self-driving laboratories, Raman data banks, and machine learning for uncovering biological insights.

Detection of microplastics via a confocal-microscope spatial-heterodyne Raman spectrometer with echelle gratings

Microplastic pollution has become a global environmental problem that cannot be ignored. Raman spectroscopy has been widely used for microplastics detection because it can be performed in real-time and is non-destructive. Conventional detection techniques have had weak signals and low signal-to-noise ratios (SNR). Here, an efficient and reliable detection method is demonstrated. Specifically, a confocal microscope combined with an echelle-grating spatial-heterodyne Raman spectrometer (CM-ESHRS) was constructed. The confocal microscopy and the characteristics of the echelle grating enabled high optical throughput, high SNR, high spectral resolution, and a wide spectral detection band. After spectral calibration, the resolution approached 0.67 cm-1, moreover, the spectral detection range for a single order was 1372.16 cm-1. We detected and analyzed nineteen kinds of microplastics, such as polyamide, polypropylene, and polymethylmethacrylate, and the main vibrational spectral bands were categorized. Compared with commercial dispersive spectrometers, CM-ESHRS has a higher optical throughput. In addition, we examined microplastics with various particle sizes, microplastics mixed in flour, and microplastic particles of different materials under mixed conditions, all of which yielded complete spectral information. Overall, CM-ESHRS exhibits good potential applications for the detection of microplastics.

High-sensitivity methane detection based on QEPAS and H-QEPAS technologies combined with a self-designed 8.7 kHz quartz tuning fork

Methane (CH4) is a greenhouse gas as well as being flammable and explosive. In this manuscript, quartz-enhanced photoacoustic spectroscopy (QEPAS) and heterodyne QEPAS (H-QEPAS) exploring a self-designed quartz tuning fork (QTF) with resonance frequency (f0) of ∼8.7 kHz was utilized to achieve sensitive CH4 detection. Compared with the standard commercial 32.768 kHz QTF, this self-designed QTF with a low f0 and large prong gap has the merits of long energy accumulation time and low optical noise. The strongest line located at 6057.08 cm-1 in the 2v3 overtone band of CH4 was chosen as the target absorption line. A diode laser with a high output power of > 30 mW was utilized as the excitation source. Acoustic micro-resonators (AmRs) were added to the sensor architecture to amplify the intensity of acoustic waves. Compared to the bare QTF, after the addition of AmRs, a signal enhancement of 149-fold and 165-fold were obtained for QEPAS and H-QEPAS systems, respectively. The corresponding minimum detection limits (MDLs) were 711 ppb and 1.06 ppm for QEPAS and H-QEPAS sensors. Furthermore, based on Allan variance analysis the MDLs can be improved to 19 ppb and 27 ppb correspondingly. Compared to the QEPAS sensor, the H-QEPAS sensor shows significantly shorter measurement timeframes, allowing for measuring the gas concentration quickly while simultaneously obtaining f0 of QTF.

Mid-infrared all-fiber light-induced thermoelastic spectroscopy sensor based on hollow-core anti-resonant fiber

In this article, a mid-infrared all-fiber light-induced thermoelastic spectroscopy (LITES) sensor based on a hollow-core anti-resonant fiber (HC-ARF) was reported for the first time. The HC-ARF was applied as a light transmission medium and gas chamber. The constructed all-fiber structure has merits of low loss, easy optical alignment, good system stability, reduced sensor size and cost. The mid-infrared transmission structure can be utilized to target the strongest gas absorption lines. The reversely-tapered SM1950 fiber and the HC-ARF were spatially butt-coupled with a V-shaped groove between the two fibers to facilitate gas entry. Carbon monoxide (CO) with an absorption line at 4291.50 cm-1 (2.33 µm) was chosen as the target gas to verify the sensing performance. The experimental results showed that the all-fiber LITES sensor based on HC-ARF had an excellent linear response to CO concentration. Allan deviation analysis indicated that the system had excellent long-term stability. A minimum detection limit (MDL) of 3.85 ppm can be obtained when the average time was 100 s.

RSPSSL: A novel high-fidelity Raman spectral preprocessing scheme to enhance biomedical applications and chemical resolution visualization

Raman spectroscopy has tremendous potential for material analysis with its molecular fingerprinting capability in many branches of science and technology. It is also an emerging omics technique for metabolic profiling to shape precision medicine. However, precisely attributing vibration peaks coupled with specific environmental, instrumental, and specimen noise is problematic. Intelligent Raman spectral preprocessing to remove statistical bias noise and sample-related errors should provide a powerful tool for valuable information extraction. Here, we propose a novel Raman spectral preprocessing scheme based on self-supervised learning (RSPSSL) with high capacity and spectral fidelity. It can preprocess arbitrary Raman spectra without further training at a speed of ~1 900 spectra per second without human interference. The experimental data preprocessing trial demonstrated its excellent capacity and signal fidelity with an 88% reduction in root mean square error and a 60% reduction in infinite norm ([Formula: see text]) compared to established techniques. With this advantage, it remarkably enhanced various biomedical applications with a 400% accuracy elevation (ΔAUC) in cancer diagnosis, an average 38% (few-shot) and 242% accuracy improvement in paraquat concentration prediction, and unsealed the chemical resolution of biomedical hyperspectral images, especially in the spectral fingerprint region. It precisely preprocessed various Raman spectra from different spectroscopy devices, laboratories, and diverse applications. This scheme will enable biomedical mechanism screening with the label-free volumetric molecular imaging tool on organism and disease metabolomics profiling with a scenario of high throughput, cross-device, various analyte complexity, and diverse applications.

Quartz-enhanced photoacoustic spectroscopy sensing using trapezoidal- and round-head quartz tuning forks

In this Letter, two novel, to the best of our knowledge, quartz tuning forks (QTFs) with trapezoidal-head and round-head were designed and adopted for quartz-enhanced photoacoustic spectroscopy (QEPAS) sensing. Based on finite element analysis, a theoretical simulation model was established to optimize the design of QTF. For performance comparison, a reported T-head QTF and a commercial QTF were also investigated. The designed QTFs have decreased resonant frequency (f0) and increased gap between the two prongs of QTF. The experimentally determined f0 of the T-head QTF, trapezoidal-head QTF, and round-head QTF were 8690.69 Hz, 9471.67 Hz, and 9499.28 Hz, respectively. The corresponding quality (Q) factors were measured as 11,142, 11,411, and 11,874. Compared to the commercial QTF, the resonance frequencies of these QTFs have reduced by 73.45%, 71.07%, and 70.99% while maintaining a comparable Q factor to the commercially mature QTF. Methane (CH4) was chosen as the analyte to verify the QTFs' performance. Compared with the commercial QTF, the signal-to-noise ratio (SNR) of the CH4-QEPAS system based on the T-head QTF, trapezoidal-head QTF, and round-head QTF has been improved by 1.75 times, 2.96 times, and 3.26 times, respectively. The performance of the CH4-QEPAS sensor based on the QTF with the best performance of the round-head QTF was investigated in detail. The results indicated that the CH4-QEPAS sensor based on the round-head QTF exhibited an excellent linear concentration response. Furthermore, a minimum detection limit (MDL) of 0.87 ppm can be achieved when the system's average time was 1200 s.

Current Trends of Raman Spectroscopy in Clinic Settings: Opportunities and Challenges

Early clinical diagnosis, effective intraoperative guidance, and an accurate prognosis can lead to timely and effective medical treatment. The current conventional clinical methods have several limitations. Therefore, there is a need to develop faster and more reliable clinical detection, treatment, and monitoring methods to enhance their clinical applications. Raman spectroscopy is noninvasive and provides highly specific information about the molecular structure and biochemical composition of analytes in a rapid and accurate manner. It has a wide range of applications in biomedicine, materials, and clinical settings. This review primarily focuses on the application of Raman spectroscopy in clinical medicine. The advantages and limitations of Raman spectroscopy over traditional clinical methods are discussed. In addition, the advantages of combining Raman spectroscopy with machine learning, nanoparticles, and probes are demonstrated, thereby extending its applicability to different clinical phases. Examples of the clinical applications of Raman spectroscopy over the last 3 years are also integrated. Finally, various prospective approaches based on Raman spectroscopy in clinical studies are surveyed, and current challenges are discussed.

Dual-frequency modulated heterodyne quartz-enhanced photoacoustic spectroscopy

A novel dual-frequency modulated heterodyne quartz-enhanced photoacoustic spectroscopy (DFH-QEPAS) was demonstrated for what we believe to be the first time in this study. In traditional H-QEPAS, the frequency of modulated sinusoidal wave has a frequency difference (Δf) with the resonance frequency (f0) of a quartz tuning fork (QTF). Owing to the resonance characteristic of QTF, it cannot excite QTF to the strongest response. To achieve a stronger response, a sinusoidal wave with a frequency of f0 was added to the modulation wave to compose a dual-frequency modulation. Acetylene (C2H2) was chosen as the target gas to verify the sensor performance. The proposed DFH-QEPAS improved 4.05 times of signal-to-noise ratio (SNR) compared with the traditional H-QEPAS in the same environmental conditions.

Multiscale Bioresponses of Metal Nanoclusters

Metal nanoclusters (NCs) are well-recognized novel nano-agents that hold great promise for applications in nanomedicine because of their ultrafine size, low toxicity, and high renal clearance. As foreign substances, however, an in-depth understanding of the bioresponses to metal NCs is necessary but is still far from being realized. Herein, this review is deployed to summarize the biofates of metal NCs at various biological levels, emphasizing their multiscale bioresponses at the molecular, cellular, and organismal levels. In the parts-to-whole schema, the interactions between biomolecules and metal NCs are discussed, presenting typical protein-dictated nano-bio interfaces, hierarchical structures, and in vivo trajectories. Then, the accumulation, internalization, and metabolic evolution of metal NCs in the cellular environment and as-imparted theranostic functionalization are demonstrated. The organismal metabolism and transportation processes of the metal NCs are subsequently distilled. Finally, this review ends with the conclusions and perspectives on the outstanding issues of metal NC-mediated bioresponses in the near future. This review is expected to provide inspiration for tailoring the customization of metal NC-based nano-agents to meet practical requirements in different sectors of nanomedicine.

Robust point light source calibration method for near-field photometric stereo using feature points selection

In this paper, we present a robust method for nonisotropic point light source calibration through feature points selection. By analyzing the relationship between the observed surface and its image intensity under near-field lighting, the feature points selection method is first developed to effectively address the noisy observations and improve calibration robustness. Afterward, to enhance efficiency and accuracy of the calibration, a cost function of l p-norm is established based on the above relationship, and an improved Newton method-based iteration process is applied to calculate the light source parameters. The simulations demonstrate that the proposed method is capable of achieving robust calibration results with the estimation error less than 2.7 mm and 0.8°, even though the image intensities are corrupted by Gaussian white noise with standard deviation up to 0.4. The experimental validation is performed using a self-designed photometric stereo system, where the calibration of point light sources is conducted, and measurements are taken on a standard sphere and compressor blade based on the obtained calibration results, which demonstrates the effectiveness of what we believe to be a new method.

High-precision prediction of blood glucose concentration utilizing Fourier transform Raman spectroscopy and an ensemble machine learning algorithm

Raman spectroscopy has gained popularity in analyzing blood glucose levels due to its non-invasive identification and minimal interference from water. However, the challenge lies in how to accurately predict blood glucose concentrations in human blood using Raman spectroscopy. This paper researches a novel integrated machine learning algorithm called Bagging-ABC-ELM. The optimal input weights and biases of extreme learning machine (ELM) model are obtained by artificial bee colony (ABC) algorithm. The bagging algorithm is used to obtain a better the stability of the model and higher performance than ELM algorithm. The results show that the mean value of coefficient of determination is 0.9928, and root mean square error is 0.1928. Compared to other regression models, the Bagging-ABC-ELM model exhibited superior prediction accuracy, robustness, and generalization capability. The Bagging-ABC-ELM model presents a promising alternative for analyzing blood glucose levels in clinical and research settings.

Machine learning applications for multi-source data of edible crops: A review of current trends and future prospects

The quality and safety of edible crops are key links inseparable from human health and nutrition. In the era of rapid development of artificial intelligence, using it to mine multi-source information on edible crops provides new opportunities for industrial development and market supervision of edible crops. This review comprehensively summarized the applications of multi-source data combined with machine learning in the quality evaluation of edible crops. Multi-source data can provide more comprehensive and rich information from a single data source, as it can integrate different data information. Supervised and unsupervised machine learning is applied to data analysis to achieve different requirements for the quality evaluation of edible crops. Emphasized the advantages and disadvantages of techniques and analysis methods, the problems that need to be overcome, and promising development directions were proposed. To monitor the market in real-time, the quality evaluation methods of edible crops must be innovated.

Towards Raman imaging of centimeter scale tissue areas for real-time opto-molecular visualization of tissue boundaries for clinical applications

Raman spectroscopy combined with augmented reality and mixed reality to reconstruct molecular information of tissue surface.