Non-invasive screening and subtyping for breast cancer by serum SERS combined with LGB-DNN algorithms.

Early detection of breast cancer and its molecular subtyping is crucial for guiding clinical treatment and improving survival rate. Current diagnostic methods for breast cancer are invasive, time consuming and complicated. In this work, an optical detection method integrating surface-enhanced Raman spectroscopy (SERS) technology with feature selection and deep learning algorithm was developed for identifying serum components and building diagnostic model, with the aim of efficient and accurate noninvasive screening of breast cancer. First, the high quality of serum SERS spectra from breast cancer (BC), breast benign disease (BBD) patients and healthy controls (HC) were obtained. Chi-square tests were conducted to exclude confounding factors, enhancing the reliability of the study. Then, LightGBM (LGB) algorithm was used as the base model to retain useful features to significantly improve classification performance. The DNN algorithm was trained through backpropagation, adjusting the weights and biases between neurons to improve the network's predictive ability. In comparison to traditional machine learning algorithms, this method provided more accurate information for breast cancer classification, with classification accuracies of 91.38 % for BC and BBD, and 96.40 % for BC, BBD, and HC. Furthermore, the accuracies of 90.11 % for HR+/HR- and 88.89 % for HER2+/HER2- can be reached when evaluating BC patients' molecular subtypes. These results demonstrate that serum SERS combined with powerful LGB-DNN algorithm would provide a supplementary method for clinical breast cancer screening.

Optical Nanobiosensor Based on Surface-Enhanced Raman Spectroscopy and Catalytic Hairpin Assembly for Early-Stage Lung Cancer Detection via Blood Circular RNA.

Lung cancer has become the leading cause of cancer-related deaths globally. However, early detection of lung cancer remains challenging, resulting in poor outcomes for the patients. Herein, we developed an optical biosensor integrating surface-enhanced Raman spectroscopy (SERS) with a catalyzed hairpin assembly (CHA) to detect circular RNA (circRNA) associated with tumor formation and progression (circSATB2). The signals of the Raman reporter were considerably enhanced by generating abundant SERS "hot spots" with a core-shell nanoprobe and 2D SERS substrate with calibration capabilities. This approach enabled the sensitive (limit of detection: 0.766 fM) and reliable quantitative detection of the target circRNA. Further, we used the developed biosensor to detect the circRNA in human serum samples, revealing that patients with lung cancer had higher circRNA concentrations than healthy subjects. Moreover, we characterized the unique circRNA concentration profiles of the early stages (IA and IB) and subtypes (IA1, IA2, and IA3) of lung cancer. These results demonstrate the potential of the proposed optical sensing nanoplatform as a liquid biopsy and prognostic tool for the early screening of lung cancer.

An outlier removal method based on PCA-DBSCAN for blood-SERS data analysis.

Surface-enhanced Raman spectroscopy (SERS) has shown promising potential in cancer screening. In practical applications, Raman spectra are often affected by deviations from the spectrometer, changes in measurement environments, and anomalies in spectrum characteristic peak intensities due to improper sample storage. Previous research has overlooked the presence of outliers in categorical data, leading to significant impacts on model learning outcomes. In this study, we propose a novel method, called Principal Component Analysis and Density Based Spatial Clustering of Applications with Noise (PCA-DBSCAN) to effectively remove outliers. This method employs dimensionality reduction and spectral data clustering to identify and remove outliers. The PCA-DBSCAN method introduces adjustable parameters (Eps and MinPts) to control the clustering effect. The effectiveness of the proposed PCA-DBSCAN method is verified through modeling on outlier-removed datasets. Further refinement of the machine learning model and PCA-DBSCAN parameters resulted in the best cancer screening model, achieving 97.41% macro-average recall and 97.74% macro-average F1-score. This paper introduces a new outlier removal method that significantly improves the performance of the SERS cancer screening model. Moreover, the proposed method serves as inspiration for outlier detection in other fields, such as biomedical research, environmental monitoring, manufacturing, quality control, and hazard prediction.

Laser tweezer Raman spectroscopy combined with deep neural networks for identification of liver cancer cells.

Rapid identification of cancer cells is crucial for clinical treatment guidance. Laser tweezer Raman spectroscopy (LTRS) that provides biochemical characteristics of cells can be used to identify cell phenotypes through classification models in a non-invasive and label-free manner. However, traditional classification methods require extensive reference databases and clinical experience, which is challenging when sampling at inaccessible locations. Here, we describe a classification method combing LTRS with deep neural network (DNN) for differential and discriminative analysis of multiple liver cancer (LC) cells. By using LTRS, we obtained high-quality single-cell Raman spectra of normal hepatocytes (HL-7702) and liver cancer cell lines (SMMC-7721, Hep3B, HepG2, SK-Hep1 and Huh7). The tentative assignment of Raman peaks indicated that arginine content was elevated and phenylalanine, glutathione and glutamate content was decreased in liver cancer cells. Subsequently, we randomly selected 300 spectra from each cell line for DNN model analysis, achieving a mean accuracy of 99.2%, a mean sensitivity of 99.2% and a mean specificity of 99.8% for the identification and classification of multiple LC cells and hepatocyte cells. These results demonstrate the combination of LTRS and DNN is a promising method for rapid and accurate cancer cell identification at single cell level.

Machine learning-assisted global DNA methylation fingerprint analysis for differentiating early-stage lung cancer from benign lung diseases.

DNA methylation plays a critical role in the development of human tumors. However, routine characterization of DNA methylation can be time-consuming and labor-intensive. We herein describe a sensitive, simple surface-enhanced Raman spectroscopy (SERS) approach for identifying the DNA methylation pattern in early-stage lung cancer (LC) patients. By comparing SERS spectra of methylated DNA bases or sequences with their counterparts, we identified a reliable spectral marker of cytosine methylation. To move toward clinical applications, we applied our SERS strategy to detect the methylation patterns of genomic DNA (gDNA) extracted from cell line models as well as formalin-fixed paraffin-embedded tissues of early-stage LC and benign lung diseases (BLD) patients. In a clinical cohort of 106 individuals, our results showed distinct methylation patterns in gDNA between early-stage LC (n = 65) and BLD patients (n = 41), suggesting cancer-induced DNA methylation alterations. Combined with partial least square discriminant analysis, early-stage LC and BLD patients were differentiated with an area under the curve (AUC) value of 0.85. We believe that the SERS profiling of DNA methylation alterations, together with machine learning could potentially offer a promising new route toward the early detection of LC.

Ultrasensitive detection of SARS-CoV-2 antigen using surface-enhanced Raman spectroscopy-based lateral flow immunosensor.

The fast spread and transmission of the coronavirus 2019 (COVID-19) has become one of serious global public health problems. Herein, a surface enhanced Raman spectroscopy-based lateral flow immunoassay (LFA) was developed for the detection of SARS-CoV-2 antigen. Using uniquely designed core-shell nanoparticle with embedded Raman probe molecules as the indicator to reveal the concentration of target protein, excellent quantitative performance with a limit of detection (LOD) of 0.03 ng/mL and detection range of 10-1000 ng/mL can be achieved within 15 min. Besides, the detection of spiked virus protein in human saliva was also performed with a portable Raman spectrometer, proposing the feasibility of the method in practical applications. This easy-to-use, rapid and accurate method would provide a point-of-care testing way as the ideal alternative for current detection requirement of virus-related biomarkers.

Ratiometric SERS-based assay with "sandwich" structure for detection of serotonin.

A ratiometric nanoensemble-functionalized Surface-Enhanced Raman Spectroscopy (SERS) chip is proposed and an ultrasensitive "sandwich" structure introduced for the detection of 5-HT to achieve early diagnosis of colon cancer. The SERS-based chip contains core-shell SERS active substrates coded by different Raman tags with Raman-silent region peaks (Au@EBP@Au NR arrays and Au@MBN@Ag NPs) and then identify-function molecule modification to construct the "sandwich" structure (Au@EBP@Au NR arrays/5-HT/Au@MBN@Ag NPs). Au@EBP@Au NR arrays showed excellent SERS performance, including good uniformity with an RSD of 5.53% and an enhancement factor (EF) of 2.13 × 107. The intensity ratio of the peaks in the Raman silent region was proportional to the concentration of 5-HT in the range 5 × 10-7-1 × 10-3 M, with a detection limit (LOD) of 4.9 × 10-9 M. Excellent assay accuracy was also demonstrated, with recoveries in the range 96.80% to 104.96%. Finally, we found that 5-HT expression levels in normal human sera were much lower than those in colon cancer patients by using a SERS-based chip for determination of the concentration of 5-HT in clinical colon cancer serum. This result suggested that the proposed approach has potential for detecting 5-HT by ratiometric SERS-based chips for early diagnosis of colon cancer.

Shifted-excitation Raman difference spectroscopy for improving in vivo detection of nasopharyngeal carcinoma.

A strong fluorescence background is one of the common interference factors of Raman spectroscopic analysis in biological tissue. This study developed an endoscopic shifted-excitation Raman difference spectroscopy (SERDS) system for real-time in vivo detection of nasopharyngeal carcinoma (NPC) for the first time. Owing to the use of the SERDS method, the high-quality Raman signals of nasopharyngeal tissue could be well extracted and characterized from the complex raw spectra by removing the fluorescence interference signals. Significant spectral differences relating to proteins, phospholipids, glucose, and DNA were found between 42 NPC and 42 normal tissue sites. Using linear discriminant analysis, the diagnostic accuracy of SERDS for NPC detection was 100%, which was much higher than that of raw Raman spectroscopy (75.0%), showing the great potential of SERDS for improving the accurate in vivo detection of NPC.

Profiling of Tumor Cell-Delivered Exosome by Surface Enhanced Raman Spectroscopy-Based Biosensor for Evaluation of Nasopharyngeal Cancer Radioresistance.

Although the advancement of radiotherapy significantly improves the survival of nasopharyngeal cancer (NPC), radioresistance associated with recurrence and poor outcomes still remains a daunting challenge in the clinical scenario. Currently, effective biomarkers and convenient detection methods for predicting radioresistance have not been well established. Here, the surface-enhanced Raman spectroscopy combined with proteomics is used to firstly profile the characteristic spectral patterns of exosomes secreted from self-established NPC radioresistance cells, and reveals specific variations of proteins expression during radioresistance formation, including collagen alpha-2 (I) chain (COL1A2) that is associated with a favorable prognosis in NPC and is negatively associated with DNA repair scores and DNA repair-related genes via bioinformatic analysis. Furthermore, deep learning model-based diagnostic model is generated to accurately identify the exosomes from radioresistance group. This work demonstrates the promising potential of exosomes as a novel biomarker for predicting the radioresistance and develops a rapid and sensitive liquid biopsy method that will provide a personalized and precise strategy for clinical NPC treatment.

Self-calibration SERS sensor with "core-satellite" structure for detection of hyaluronidase activity.

Hyaluronidase expression is known to be upregulated in various pathological conditions. A method based on a combination of ratiometric surface-enhanced Raman scattering (SERS) and magnetic separation is described for the determination of hyaluronidase (HAase) activity. Gold nanospheres (AuNPs) functionalized by 4-mercaptophenylboronic acid (4-MPBA) form stable cyclic esters with diol on hyaluronic acid (HA) by the boronic acid group, while Fe3O4@DTNB@Au modified with mercaptoethylamine (MEA) was used as a capture substrate to bind to the carboxyl group on the surface of HA, forming the "Au@4-MPBA@HA/Fe3O4@DTNB@Au@MEA" "core-satellite" structure. When HAase is present, HA is enzymatically disrupted, resulting in the destruction of the "core-satellite" structure, the SERS signal of 4-MPBA is subsequently weakened. The gold shell in the substrate protects the 5,5'-Dithio bis-(2-nitrobenzoic acid) (DTNB) from the external environment, which makes it become an ideal internal standard (IS) molecule for subsequent calibration. Under optimal conditions, the I1075/I1324 varied in the range of 10-3 - 10 U‧mL-1 HAase activity, with a limit of detection (LOD) of 0.32 mU‧mL-1，below the level of HAase in normal human body fluids. This method has been successfully applied to the determination of HAase activity in urine and is expected to provide a new method in disease detection, especially in the non-invasive detection of bladder cancer.

Highly Efficient Blood Protein Analysis Using Membrane Purification Technique and Super-Hydrophobic SERS Platform for Precise Screening and Staging of Nasopharyngeal Carcinoma.

Early screening and precise staging are crucial for reducing mortality in patients with nasopharyngeal carcinoma (NPC). This study aimed to assess the performance of blood protein surface-enhanced Raman scattering (SERS) spectroscopy, combined with deep learning, for the precise detection of NPC. A highly efficient protein SERS analysis, based on a membrane purification technique and super-hydrophobic platform, was developed and applied to blood samples from 1164 subjects, including 225 healthy volunteers, 120 stage I, 249 stage II, 291 stage III, and 279 stage IV NPC patients. The proteins were rapidly purified from only 10 µL of blood plasma using the membrane purification technique. Then, the super-hydrophobic platform was prepared to pre-concentrate tiny amounts of proteins by forming a uniform deposition to provide repeatable SERS spectra. A total of 1164 high-quality protein SERS spectra were rapidly collected using a self-developed macro-Raman system. A convolutional neural network-based deep-learning algorithm was used to classify the spectra. An accuracy of 100% was achieved for distinguishing between the healthy and NPC groups, and accuracies of 96%, 96%, 100%, and 100% were found for the differential classification among the four NPC stages. This study demonstrated the great promise of SERS- and deep-learning-based blood protein testing for rapid, non-invasive, and precise screening and staging of NPC.

Ratiometric SERS quantitative analysis of tyrosinase activity based on gold-gold hybrid nanoparticles with Prussian blue as an internal standard.

Tyrosinase (TYR) is a polyphenol oxidase that regulates melanin biosynthesis. Abnormal levels of TYR have been confirmed closely associated with melanoma cancer and other diseases, making the establishment of highly sensitive and accurate quantitative detection of TYR is thus essential for fundamental research and clinical applications. Herein, we proposed a new strategy that combines surface-enhanced Raman scattering (SERS) with Dopamine (DA) and Prussian blue (PB) functionalized gold-gold hybrid nanoparticles for TYR detection. DA is oxidized to dopaquinone with the presence of TYR, leading to the consumption of DA in the reaction solution and the corresponding decrease in DA characteristic peak intensity at 1480 cm-1. Our SERS quantitative assay of TYR with "on-off" sensing strategy yields an excellent limit of detection (LOD) of 3 × 10-3 U mL-1 in a linear range of 10-3 to 100 U mL-1. In particular, the intensity ratio of 1480 cm-1 to 2121 cm-1 allows us to achieve remarkably reliable quantitative detection of TYR, with the 2121 cm-1 peak intensity of the standard internal (IS) molecule PB being used to correct SERS signal fluctuations. Furthermore, our proposed assay has been successfully demonstrated to quantify TYR spiked in human serum samples, with excellent percentage recovery of 90.0-110.6 %. The potential of our ratiometric SERS strategy for the performance evaluation of TYR inhibitors has also been verified. Our work is therefore expected to provide a new route for accurate and reliable monitoring of TYR activity in the complex biological environment.

Differentiation and prognostic stratification of acute myeloid leukemia by serum-based spectroscopy coupling with metabolic fingerprints.

Acute myeloid leukemia (AML) is a heterogeneous disease characterized by complex molecular and cytogenetic abnormalities. New approaches to predict the prognosis of AML have increasingly attracted attention. There were 98 non-M3 AML cases and 48 healthy controls were enrolled in the current work. Clinically routine assays for cytogenetic and molecular genetic analyses were performed on the bone marrow samples of patients with AML. Meanwhile, metabolic profiling of these AML subjects was also performed on the serum samples by combining Ag nanoparticle-based surface-enhanced Raman spectroscopy (SERS) with proton nuclear magnetic resonance (NMR) spectroscopy. Although most of the routine biochemical test showed no significant differences between the M0-M2 and M5 groups, the metabolic profiles were significantly different either between AML subtypes or between prognostic risk subgroups. Specific SERS bands were screened to serve as potential markers for AML subtypes. The results demonstrated that the classification models for M0-M2 and M5 shared two bands (i.e., 1328 and 741 cm-1 ), all came from nucleic acid signals. Furthermore, Metabolic profiles provided various differential metabolites responsible for different AML subtypes, and we found altered pathways mainly included energy metabolism like glycolysis, pyruvate metabolism, and metabolisms of nucleic acid bases as well as specific amino acid metabolisms. It is concluded that integration of SERS and NMR provides the rational and could be reliable to reveal AML differentiation, and meanwhile lay the basis for experimental and clinical practice to monitor disease progression and prognostic evaluation.

Early detection of lung cancer via biointerference-free, target microRNA-triggered core-satellite nanocomposites.

MicroRNAs (miRNAs) are emerging as essential liquid biopsy markers for early cancer detection. Currently, the clinical applications of miRNAs are lagging behind due to their high sequence similarity and rarity. Herein, we propose biointerference-free, target-triggered core-satellite nanocomposites for ultrasensitive surface-enhanced Raman spectroscopy (SERS) detection of lung cancer-related miRNA-21. Through the hybridization-based recognition effect, we observe an enormous SERS signal enhancement caused by miRNA-21-triggered assembly of core-satellite nanocomposites. This enables the sensitive detection of miRNA-21 down to the 0.1 fM level in a linear range of 10 fM to 1 nM. The use of a biointerference-free reporter further allows quantitative and direct detection of miRNA-21 from complex plasma samples, without RNA pre-extraction. As a proof of principle, we measure the level of plasma miRNA-21 in 20 lung cancer patients and 10 healthy participants. Significantly higher levels of miRNA-21 are determined in lung cancer patients than in healthy participants, with clear lower expression in stage I (n = 10) than in stage III-IV (n = 10) lung cancer patients. We, therefore, believe that this proposed strategy will have high clinical potential for sensitive quantification of miRNA markers in liquid biopsy samples and act as a complementary method for the early detection of lung cancer.

Highly sensitive and reliable detection of microRNA for clinically disease surveillance using SERS biosensor integrated with catalytic hairpin assembly amplification technology.

MicroRNAs (miRNAs) play an important regulatory role in several diseases, especially as a class of promising biomarkers for cancer diagnosis and prognosis. Here, a biosensor based on surface enhanced Raman spectroscopy (SERS) combined with catalytic hairpin assembly (CHA) amplification technology was developed for ultra-sensitive detection of miRNA-21 and miRNA-155 in breast cancer serum. By using CHA strategy, the extremely low concentration of target microRNA in human serum can be significantly amplified through the re-hybridization with thousands of hairpin probes to trigger amplification cycles. Besides, a sandwich SERS sensing chip with numerous hot spots and signal self-calibration was built through the linkage between two-dimensional Au-Si substrate and upper Ag@4-MBA@Au core-shell nanoparticles. Using this specially-designed biosensing platform, a low detection limit of 0.398 fM and 0.215 fM with a dynamic range from 1 fM to 10 nM can be achieved for the detection of miRNA-21 and miRNA-155, respectively. Additionally, the analysis of these two miRNAs in serum samples is capable of identifying the breast cancer subjects from normal ones with 100% of accuracy, as well as potentially evaluating the molecular types and prognosis for breast cancer. These results demonstrate that the proposed SERS with CHA technology would be an alternative method for highly sensitive and reliable detection of miRNA biomarkers contributing to breast cancer diagnosis and prognosis.

An Endoscope-like SERS Probe Based on the Focusing Effect of Silica Nanospheres for Tyrosine and Urea Detection in Sweat.

In this work, we developed a new type of SERS probe, which was composed of glass-SiO2-Au@MBN@Ag nanoparticles (NPs) three-dimensional Surface-enhanced Raman spectroscopy (SERS) substrate. When the laser passed through the quartz glass sheet, on the one hand, the SiO2 NPs supporting the Au@MBN@Ag NPs increase the roughness of the substrate surface, resulting in a large number of hot spots among nanoparticles. On the other hand, based on the focusing effect of silicon dioxide nanospheres, the laser can better focus on the surface of nanoparticles in the inverted SERS probe, thus showing better SERS enhancement. Furthermore, the Au@MBN@Ag NPs core-shell structure was used with 4-mercaptobenzoonitrile (MBN) as an internal standard molecule, and the quantitative determination of tyrosine and urea was realized by internal standard correction method. The standard working curves of the two had good linear correlation with R2 above 0.9555. The detection limits of tyrosine and urea were in the range of 2.85 × 10-10 M~7.54 × 10-6 M, which confirms that this design can be used for quantitative and specific detection of biological molecules, demonstrating great practical significance for the research of diseases such as skin lesions and endocrine disorders.

Subtype discrimination of acute myeloid leukemia based on plasma SERS technique.

Acute myeloid leukemia (AML) is a common hematologic malignancy. To this day, diagnose of AML and its genetic mutation still rely on invasive and time-consuming methods. In this study, 222 plasma samples were collected to discuss the performance of surface-enhanced Raman spectroscopy (SERS) to discriminate AML subtype acute promyelocytic leukemia and acute monocytic leukemia based on plasma. The Ag nanoparticles-based SERS technique was used to explore the biochemical differences among different AML subtypes. With the help of powerful supervised and unsupervised algorithms, the performance using the whole spectra and band intensities was confirmed to identify different subtypes of AML. The results demonstrated the intensities of several bands and band-intensity ratios were significantly different between groups, thus related to the discrimination of several AML subtypes and control. Combining indexes of band-intensity ratios, the result of multi-indexes ROC has excellent performance in differentiating AML patient with healthy control. Our work demonstrated the great potential of SERS technique as a rapid and micro detection method in clinical laboratory field, it's a new and powerful tool for analyzing human blood plasma.

Optical biosensor based on SERS with signal calibration function for quantitative detection of carcinoembryonic antigen.

Monitoring the levels of cancer biomarkers is essential for cancer diagnosis and evaluation. In this study, a novel sandwich type sensing platform based on surface-enhanced Raman scattering (SERS) technology was developed for the detection of carcinoembryonic antigen (CEA), with a limit of detection (LOD) of 0.258 ng/mL. In order to achieve sensitive detection of CEA in complex samples, gold nanoparticle monolayer modified with CEA antibodies and with aptamer-functionalized probes was fabricated to target CEA. Two gold layers were integrated into the SERS platform, which greatly enhanced the signal of the probe by generating tremendous "hot spots". Meanwhile, the intensity ratio of Raman probes and the second-order peak of the silicon wafer was used to achieve dynamic calibration of the Raman probe signal. Excitingly, this sensing platform was capable of distinguishing cancer patients from healthy individuals via CEA concentrations in blood samples with the accuracy of 100%. This sandwich structure SERS sensing platform presented promising potential to be an alternative tool for clinical biomarker detection in the field of cancer diagnosis.

Characterization of Drug Resistance in Chronic Myeloid Leukemia Cells Based on Laser Tweezers Raman Spectroscopy.

Multidrug resistance is highly associated with poor prognosis of chronic myeloid leukemia. This work aims to explore whether the laser tweezers Raman spectroscopy (LTRS) could be practical in separating adriamycin-resistant chronic myeloid leukemia cells K562/adriamycin from its parental cells K562, and to explore the potential mechanisms. Detection of LTRS initially reflected the spectral differences caused by chemoresistance including bands assigned to carbohydrates, amino acid, protein, lipids, and nucleic acid. In addition, principal components analysis as well as the classification and regression trees algorithms showed that the specificity and sensitivity were above 90%. Moreover, the band data-based classification and regression tree model and receiver operating characteristic curve further determined some important bands and band intensity ratios to be reliable indexes in discriminating K562 chemoresistance status. Finally, we highlighted three metabolism pathways correlated with chemoresistance. This work demonstrates that the label-free LTRS analysis combined with multivariate statistical analyses have great potential to be a novel analytical strategy at the single-cell level for rapid evaluation of the chemoresistance status of K562 cells.

Erratum: Assessment of radiotherapy effect for nasopharyngeal cancer using plasma surface-enhanced Raman spectroscopy technology: errata.

[This corrects the article on p. 3413 in vol. 9, PMID: 29984106.].