ISREA: An Efficient Peak-Preserving Baseline Correction Algorithm for Raman Spectra

A Study of High-Frequency Noise for Microplastics Classification Using Raman Spectroscopy and Machine Learning

Given the growing urge for plastic management and regulation in the world, recent studies have investigated the problem of plastic material identification for correct classification and disposal. Recent works have shown the potential of machine learning techniques for successful microplastics classification using Raman signals. Classification techniques from the machine learning area allow the identification of the type of microplastic from optical signals based on Raman spectroscopy. In this paper, we investigate the impact of high-frequency noise on the performance of related classification tasks. It is well-known that classification based on Raman is highly dependent on peak visibility, but it is also known that signal smoothing is a common step in the pre-processing of the measured signals. This raises a potential trade-off between high-frequency noise and peak preservation that depends on user-defined parameters. The results obtained in this work suggest that a linear discriminant analysis model cannot generalize properly in the presence of noisy signals, whereas an error-correcting output codes model is better suited to account for inherent noise. Moreover, principal components analysis (PCA) can become a must-do step for robust classification models, given its simplicity and natural smoothing capabilities. Our study on the high-frequency noise, the possible trade-off between pre-processing the high-frequency noise and the peak visibility, and the use of PCA as a noise reduction technique in addition to its dimensionality reduction functionality are the fundamental aspects of this work.

Cancer detection in dogs using rapid Raman molecular urinalysis

Introduction

The presence of cancer in dogs was detected by Raman spectroscopy of urine samples and chemometric analysis of spectroscopic data. The procedure created a multimolecular spectral fingerprint with hundreds of features related directly to the chemical composition of the urine specimen. These were then used to detect the broad presence of cancer in dog urine as well as the specific presence of lymphoma, urothelial carcinoma, osteosarcoma, and mast cell tumor.

Methods

Urine samples were collected via voiding, cystocentesis, or catheterization from 89 dogs with no history or evidence of neoplastic disease, 100 dogs diagnosed with cancer, and 16 dogs diagnosed with non-neoplastic urinary tract or renal disease. Raman spectra were obtained of the unprocessed bulk liquid urine samples and were analyzed by ISREA, principal component analysis (PCA), and discriminant analysis of principal components (DAPC) were applied using the Rametrix®Toolbox software.

Results and discussion

The procedure identified a spectral fingerprint for cancer in canine urine, resulting in a urine screening test with 92.7% overall accuracy for a cancer vs. cancer-free designation. The urine screen performed with 94.0% sensitivity, 90.5% specificity, 94.5% positive predictive value (PPV), 89.6% negative predictive value (NPV), 9.9 positive likelihood ratio (LR+), and 0.067 negative likelihood ratio (LR-). Raman bands responsible for discerning cancer were extracted from the analysis and biomolecular associations were obtained. The urine screen was more effective in distinguishing urothelial carcinoma from the other cancers mentioned above. Detection and classification of cancer in dogs using a simple, non-invasive, rapid urine screen (as compared to liquid biopsies using peripheral blood samples) is a critical advancement in case management and treatment, especially in breeds predisposed to specific types of cancer.

Key steps for improving bacterial SERS signals in complex samples: Separation, recognition, detection, and analysis

Rapid and reliable detection of pathogenic bacteria is absolutely essential for research in environmental science, food quality, and medical diagnostics. Surface-enhanced Raman spectroscopy (SERS), as an emerging spectroscopic technique, has the advantages of high sensitivity, good selectivity, rapid detection speed, and portable operation, which has been broadly used in the detection of pathogenic bacteria in different kinds of complex samples. However, the SERS detection method is also challenging in dealing with the detection difficulties of bacterial samples in complex matrices, such as interference from complex matrices, confusion of similar bacteria, and complexity of data processing. Therefore, researchers have developed some technologies to assist in SERS detection of bacteria, including both the front-end process of obtaining bacterial sample data and the back-end data processing process. The review summarizes the key steps for improving bacterial SERS signals in complex samples: separation, recognition, detection, and analysis, highlighting the principles of each step and the key roles for SERS pathogenic bacteria analysis, and the interconnectivity between each step. In addition, the current challenges in the practical application of SERS technology and the development trends are discussed. The purpose of this review is to deepen researchers' understanding of the various stages of using SERS technology to detect bacteria in complex sample matrices, and help them find new breakthroughs in different stages to facilitate the detection and control of bacteria in complex samples.

Enhanced 3-D asynchronous correlation data preprocessing method for Raman spectroscopy of Chinese handmade paper

We have developed a novel 3D asynchronous correlation method (3D-ACM) designed for the classification and identification of Chinese handmade paper samples using Raman spectra and machine learning. The 3D-ACM approach involves two rounds of tensor product and Hilbert transform operations. In the tensor product process, the outer product of the spectral data from different samples within the same category is computed, establishing inner connections among all samples within that category. The Hilbert transform introduces a 90-degree phase shift, resulting in a true three-dimensional spectral data structure. This expansion significantly increases the number of equivalent frequency points and samples within each category. This enhancement substantially boosts spectral resolution and reveals more hidden information within the spectral data. To maximize the potential of 3D-ACM, we employed six machine learning models: principal component analysis (PCA) with linear regression (LR), support vector machine (SVM) with LR, k-Nearest Neighbors (KNN), random forest (RF), and convolutional neural network (CNN). When applied to the 3D-ACM data preprocessing method, R-squared values of PLS-LR, KNN, RF and CNN supervised models, approached or equaled 1. This indicates exceptional performance comparable to unsupervised models like PCA. 3D-ACM stands as a versatile mathematical technique not confined to spectral data. It also eliminates the necessity for additional experimental setups or external control conditions, distinct from traditional two-dimensional correlation spectroscopy. Moreover, it preserves the original experimental data, setting it apart from conventional data preprocessing methods. This positions 3D-ACM as a promising tool for future material classification and identification in conjunction with machine learning.

Deep neural network: As the novel pipelines in multiple preprocessing for Raman spectroscopy

Raman spectroscopy is a kind of vibrational method that can rapidly and non-invasively gives chemical structural information with the Raman spectrometer. Despite its technical advantages, in practical application scenarios, Raman spectroscopy often suffers from interference, such as noises and baseline drifts, resulting in the inability to acquire high-quality Raman spectroscopy signals, which brings challenges to subsequent spectral analysis. The commonly applied spectral preprocessing methods, such as Savitzky-Golay smooth and wavelet transform, can only perform corresponding single-item processing and require manual intervention to carry out a series of tedious trial parameters. Especially, each scheme can only be used for a specific data set. In recent years, the development of deep neural networks has provided new solutions for intelligent preprocessing of spectral data. In this paper, we first creatively started from the basic mechanism of spectral signal generation and constructed a mathematical model of the Raman spectral signal. By counting the noise parameters of the real system, we generated a simulation dataset close to the output of the real system, which alleviated the dependence on data during deep learning training. Due to the powerful nonlinear fitting ability of the neural network, fully connected network model is constructed to complete the baseline estimation task simply and quickly. Then building the Unet model can effectively achieve spectral denoising, and combining it with baseline estimation can realize intelligent joint processing. Through the simulation dataset experiment, it is proved that compared with the classic method, the method proposed in this paper has obvious advantages, which can effectively improve the signal quality and further ensure the accuracy of the peak intensity. At the same time, when the proposed method is applied to the actual system, it also achieves excellent performance compared with the common method, which indirectly indicates the effectiveness of the Raman signal simulation model. The research presented in this paper offers a variety of efficient pipelines for the intelligent processing of Raman spectroscopy, which can adapt to the requirements of different tasks while providing a new idea for enhancing the quality of Raman spectroscopy signals.

A double sliding-window method for baseline correction and noise estimation for Raman spectra of microplastics

When measuring microplastics of environmental samples, additives and attachment of biological materials may result in strong fluorescence in Raman spectra, which increases difficulty for imaging, identification, and quantification. Although there are several baseline correction methods available, user intervention is usually needed, which is not feasible for automated processes. In current study, a double sliding-window (DSW) method was proposed to estimate the baseline and standard deviation of noise. Simulated spectra and experimental spectra were used to evaluate the performance in comparison with two popular and widely used methods. Validation with simulated spectra and spectra of environmental samples showed that DSW method can accurately estimate the standard deviation of spectral noise. DSW method also showed better performance than compared methods when handling spectra of low signal-to-noise ratio (SNR) and elevated baselines. Therefore, DSW method is a useful approach for preprocessing Raman spectra of environmental samples and automated processes.

Analysis of urine Raman spectra differences from patients with diabetes mellitus and renal pathologies

Background

Chronic kidney disease (CKD) poses a major public health burden. Diabetes mellitus (DM) is one of the major causes of CKD. In patients with DM, it can be difficult to differentiate diabetic kidney disease (DKD) from other causes of glomerular damage; it should not be assumed that all DM patients with decreased eGFR and/or proteinuria have DKD. Renal biopsy is the standard for definitive diagnosis, but other less invasive methods may provide clinical benefit. As previously reported, Raman spectroscopy of CKD patient urine with statistical and chemometric modeling may provide a novel, non-invasive methodology for discriminating between renal pathologies.

Methods

Urine samples were collected from renal biopsied and non-biopsied patients presenting with CKD secondary to DM and non-diabetic kidney disease. Samples were analyzed by Raman spectroscopy, baselined with the ISREA algorithm, and subjected to chemometric modeling. Leave-one-out cross-validation was used to assess the predictive capabilities of the model.

Results

This proof-of-concept study consisted of 263 samples, including renal biopsied, non-biopsied diabetic and non-diabetic CKD patients, healthy volunteers, and the Surine™ urinalysis control. Urine samples of DKD patients and those with immune-mediated nephropathy (IMN) were distinguished from one another with 82% sensitivity, specificity, positive-predictive value (PPV), and negative-predictive value (NPV). Among urine samples from all biopsied CKD patients, renal neoplasia was identified in urine with 100% sensitivity, specificity, PPV, and NPV, and membranous nephropathy was identified with 66.7% sensitivity, 96.4% specificity, 80.0% PPV, and 93.1% NPV. Finally, DKD was identified among a population of 150 patient urine samples containing biopsy-confirmed DKD, other biopsy-confirmed glomerular pathologies, un-biopsied non-diabetic CKD patients (no DKD), healthy volunteers, and Surine™ with 36.4% sensitivity, 97.8% specificity, 57.1% PPV, and 95.1% NPV. The model was used to screen un-biopsied diabetic CKD patients and identified DKD in more than 8% of this population. IMN in diabetic patients was identified among a similarly sized and diverse population with 83.3% sensitivity, 97.7% specificity, 62.5% PPV, and 99.2% NPV. Finally, IMN in non-diabetic patients was identified with 50.0% sensitivity, 99.4% specificity, 75.0% PPV, and 98.3% NPV.

Conclusions

Raman spectroscopy of urine with chemometric analysis may be able to differentiate between DKD, IMN, and other glomerular diseases. Future work will further characterize CKD stages and glomerular pathology, while assessing and controlling for differences in factors such as comorbidities, disease severity, and other lab parameters.

Chemometrics for Raman Spectroscopy Harmonization

Raman spectroscopy is used in a wide variety of fields, and in a plethora of different configurations. Raman spectra of simple analytes can often be analyzed using univariate approaches and interpreted in a straightforward manner. For more complex spetral data such as time series or line profiles (1D), Raman maps (2D), or even volumes (3D), multivariate data analysis (MVDA) becomes a requirement. Even though there are some existing standards for creation, implementation, and validation of methods and models employed in industry and academics, further research and development in the field must contribute to their improvement. This review will cover, in broad terms, existing techniques as well as new developments for MVDA for Raman spectroscopic data, and in particular the use associated with instrumentation and data calibration. Chemometric models are often generated via fusion of analytical data from different sources, which enhances model discrimination and prediction abilities as compared to models derived from a single data source. For Raman spectroscopy, raw or unprocessed data is rarely ever used. Instead, spectra are usually corrected and manipulated,¹ often by case-specific rather than universal methods. Calibration models can be used to characterize qualitatively and/or quantitatively samples measured with the same instrumentation that was used to create the model. However, regular validation is required to ensure that aging or incorrect maintenance of the instrument does not alter the model's predictions, particularly when applied in regulated fields such as pharmaceuticals. Furthermore, a model transfer may be required for different reasons, such as replacement or significant repair of the instrumentation. Modeling can also be used to consistently harmonize Raman spectroscopic data across several instrumental designs, accounting for variations in the resulting spectrum induced by different components. Data for Raman harmonization models should be processed in a protocolled manner, and the original data accessible to allow for model reconstruction or transfer when new data is added. Important processing steps will be the calibration of the spectral axes and instrument dependent effects, such as spectral resolution. In addition, data fusion and model transfer are essential for allowing new instrumentation to build on existing models to harmonize their own data. Ideally, an open access database would be created and maintained, for the purpose of allowing for continued harmonization of new Raman instruments using an outlined and accepted protocol.

Alterations in the molecular composition of COVID-19 patient urine, detected using Raman spectroscopic/computational analysis

We developed and tested a method to detect COVID-19 disease, using urine specimens. The technology is based on Raman spectroscopy and computational analysis. It does not detect SARS-CoV-2 virus or viral components, but rather a urine ‘molecular fingerprint’, representing systemic metabolic, inflammatory, and immunologic reactions to infection. We analyzed voided urine specimens from 46 symptomatic COVID-19 patients with positive real time-polymerase chain reaction (RT-PCR) tests for infection or household contact with test-positive patients. We compared their urine Raman spectra with urine Raman spectra from healthy individuals (n = 185), peritoneal dialysis patients (n = 20), and patients with active bladder cancer (n = 17), collected between 2016–2018 (i.e., pre-COVID-19). We also compared all urine Raman spectra with urine specimens collected from healthy, fully vaccinated volunteers (n = 19) from July to September 2021. Disease severity (primarily respiratory) ranged among mild (n = 25), moderate (n = 14), and severe (n = 7). Seventy percent of patients sought evaluation within 14 days of onset. One severely affected patient was hospitalized, the remainder being managed with home/ambulatory care. Twenty patients had clinical pathology profiling. Seven of 20 patients had mildly elevated serum creatinine values (>0.9 mg/dl; range 0.9–1.34 mg/dl) and 6/7 of these patients also had estimated glomerular filtration rates (eGFR) <90 mL/min/1.73m² (range 59–84 mL/min/1.73m²). We could not determine if any of these patients had antecedent clinical pathology abnormalities. Our technology (Raman Chemometric Urinalysis—Rametrix^®) had an overall prediction accuracy of 97.6% for detecting complex, multimolecular fingerprints in urine associated with COVID-19 disease. The sensitivity of this model for detecting COVID-19 was 90.9%. The specificity was 98.8%, the positive predictive value was 93.0%, and the negative predictive value was 98.4%. In assessing severity, the method showed to be accurate in identifying symptoms as mild, moderate, or severe (random chance = 33%) based on the urine multimolecular fingerprint. Finally, a fingerprint of ‘Long COVID-19’ symptoms (defined as lasting longer than 30 days) was located in urine. Our methods were able to locate the presence of this fingerprint with 70.0% sensitivity and 98.7% specificity in leave-one-out cross-validation analysis. Further validation testing will include sampling more patients, examining correlations of disease severity and/or duration, and employing metabolomic analysis (Gas Chromatography–Mass Spectrometry [GC-MS], High Performance Liquid Chromatography [HPLC]) to identify individual components contributing to COVID-19 molecular fingerprints.

Disease-Associated Multimolecular Signature in the Urine of Patients with Lyme Disease Detected Using Raman Spectroscopy and Chemometrics

A urine-based screening technique for Lyme disease (LD) was developed in this research. The screen is based on Raman spectroscopy, iterative smoothing-splines with root error adjustment (ISREA) spectral baselining, and chemometric analysis using Rametrix software. Raman spectra of urine from 30 patients with positive serologic tests (including the US Centers for Disease Control [CDC] two-tier standard) for LD were compared against subsets of our database of urine spectra from 235 healthy human volunteers, 362 end-stage kidney disease (ESKD) patients, and 17 patients with active or remissive bladder cancer (BCA). We found statistical differences (p < 0.001) between urine scans of healthy volunteers and LD-positive patients. We also found a unique LD molecular signature in urine involving 112 Raman shifts (31 major Raman shifts) with significant differences from urine of healthy individuals. We were able to distinguish the LD molecular signature as statistically different (p < 0.001) from the molecular signatures of ESKD and BCA. When comparing LD-positive patients against healthy volunteers, the Rametrix-based urine screen performed with 86.7% for overall accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), respectively. When considering patients with ESKD and BCA in the LD-negative group, these values were 88.7% (accuracy), 83.3% (sensitivity), 91.0% (specificity), 80.7% (PPV), and 92.4% (NPV). Additional advantages to the Raman-based urine screen include that it is rapid (minutes per analysis), is minimally invasive, requires no chemical labeling, uses a low-profile, off-the-shelf spectrometer, and is inexpensive relative to other available LD tests.

Raman Spectroscopic Detection and Quantification of Macro- and Microhematuria in Human Urine

Hematuria refers to the presence of blood in urine. Even in small amounts, it may be indicative of disease, ranging from urinary tract infection to cancer. Here, Raman spectroscopy was used to detect and quantify macro- and microhematuria in human urine samples. Anticoagulated whole blood was mixed with freshly collected urine to achieve concentrations of 0, 0.25, 0.5, 1, 2, 6, 10, and 20% blood/urine (v/v). Raman spectra were obtained at 785 nm and data analyzed using chemometric methods and statistical tests with the Rametrix toolboxes for Matlab. Goldindec and iterative smoothing splines with root error adjustment (ISREA) baselining algorithms were used in processing and normalization of Raman spectra. Rametrix was used to apply principal component analysis (PCA), develop discriminate analysis of principal component (DAPC) models, and to validate these models using external leave-one-out cross-validation (LOOCV). Discriminate analysis of principal component models were capable of detecting various levels of microhematuria in unknown urine samples, with prediction accuracies of 91% (using Goldindec spectral baselining) and 94% (using ISREA baselining). Partial least squares regression (PLSR) was then used to estimate/quantify the amount of blood (v/v) in a urine sample, based on its Raman spectrum. Comparing actual and predicted (from Raman spectral computations) hematuria levels, a coefficient of determination (R²) of 0.91 was obtained over all hematuria levels (0-20% v/v), and an R² of 0.92 was obtained for microhematuria (0-1% v/v) specifically. Overall, the results of this preliminary study suggest that Raman spectroscopy and chemometric analyses can be used to detect and quantify macro- and microhematuria in unprocessed, clinically relevant urine specimens.

Baseline correction using a deep-learning model combining ResNet and UNet

A baseline correction method based on deep-learning model is proposed, which combines ResNet and UNet. Compared with the traditional methods, this method has relatively excellent performance.