As part of my thesis, I will conduct an experiment to measure glucose concentrations in blood using Raman Spectroscopy. Please see the table below for a summary of all the sources I read.

1. Rapid, Sensitive and Selective Optical Glucose Sensing with Stimulated Raman Scattering (SRS)

Title	Rapid, Sensitive and Selective Optical Glucose Sensing with Stimulated Raman Scattering (SRS)
Year	2022
Preparation	1. Calibration curve: D-(+)-glucose powder ($C_6H_{12}O_6$ , purity ≥ 99.5%) with nanopure water (concentrations of 75, 50, 25, 10, 5, and 2.5 $mol/m^3$)
	2. Sample: Human serum solution (male AB plasma, sterile-filtered, stored in 253K)
Instrument	1. The stimulated Raman spectra were obtained by direct transmission confocal microscope (Leica TCS SP8, Leica Microsystems, Germany)
	2. The Stokes beam wavelength was fixed at 1031.2 nm, and the Pump beam was tuned from 918 to 937 nm, allowing the excitation of vibrations in the range from 975 to 1200 $cm^{–1}$.
Method	SRS enhances the Raman effect by stimulating excitation, leading to improved sensitivity and selectivity in optical glucose sensing.
Preprocessing	1. Baseline correction: Asymmetric least-square fit
	2. Smoothing: Savitzky-Golay filter
Results	linear calibration curve for glucose concentrations from 2.5 to 100 $mol/m^3$ (mmol/l) with a theoretical limit of detection (LOD) of 3.5 $mol/m^3$ ($R^2$ value of 0.99.)

2. Potential of Raman spectroscopy for the analysis of plasma/serum in the liquid state: recent advances

Title	Potential of Raman spectroscopy for the analysis of plasma/serum in the liquid state: recent advances
Year	2019
Preparation	Liquid sample were prepared by 1.Centrifugation
	2.Serum Fractionation by ion exchange chromatography and mild sonication are used to separate and improve the solubility of specific components
	3. Ultrafiltration
	4. High-Throughput Sample Loading
	5. Serum Storage and Handling
Instrument	1. Raman Spectrometer
	2. Centrifugal Filtration Devices
	3. High-Throughput Substrate
	4. Inverted Microscopy Modality
	5. Water Immersion Objective
	6. Lab-Tek Plate
Method	Inverted Raman Spectral Analysis
Preprocessing	1. Baseline Correction: rubber band baseline correction, asymmetric least squares, and polynomial filters.
	2. Smoothing: The Savitzky–Golay smoothing algorithm
	3.Normalization: Vector normalization and scaling of the analyte spectra
	4. Water Subtraction: EMSC
Results	The results and findings presented in the paper demonstrate the potential of Raman spectroscopy for the rapid and accurate analysis of plasma/serum in the liquid state.

3. Measurement of Diabetic Sugar Concentration in Human Blood Using Raman Spectroscopy

Title	Measurement of Diabetic Sugar Concentration in Human Blood Using Raman Spectroscopy
Year	2012
Preparation	Liquid sample were prepared by 1.Centrifugation
	2.Serum Fractionation by ion exchange chromatography and mild sonication are used to separate and improve the solubility of specific components
	3. Ultrafiltration
	4. High-Throughput Sample Loading
	5. Serum Storage and Handling
Instrument	1.Raman spectroscopy system model MST-4000A
	2. Laser sources wavelength 532 nm and 442 nm, samples slides, chamber light collection optics detection system
Method	The measurements were performed by θ/2θ scans in the 2θ angular range of 20°–95°, with a step size of 0.02° and a scan rate of 2 deg/min
Preprocessing	Pre-processing is included in their method.
Results	The experiment indicated that Raman spectroscopy holds promise as a non-invasive method for diabetic sugar concentration in human blood, providing detailed biochemical information and diagnostic benefits in a clinical setting.

4. Simultaneous detection of glucose, triglycerides, and total cholesterol in whole blood by Fourier-Transform Raman spectroscopy

Title	Simultaneous detection of glucose, triglycerides, and total cholesterol in whole blood by Fourier-Transform Raman spectroscopy
Year	2021
Preparation	161 blood samples from Qingdao Hospital was collected by venipuncture into vacutainer blood collection tubes containing 0.2% Ethylenediaminetetraacetic acid (EDTA) as an anticoagulant.
Instrument	Raman spectra at 1064 nm excitation radiation(Nd:YAG laser with the maximum output power of 500mW) were recorded with a FT-Raman Bruker MultiRAM spectrometer (BrukerOptics, Germany)
Method	-----
Preprocessing	1. Baseline correction+PLS
	2. Orthogonal Signal Correction(OSC)
	3. Mehalanobis Distance Algorithm
	4. Sample partition: SPXY algorithm
Results	The results showed that the proposed method accurately predicted the concentration of glucose, total cholesterol (TC), and triglycerides (TG) in blood.

5. Quantitative analysis of human blood serum using vibrational spectroscopy

Title	Quantitative analysis of human blood serum using vibrational spectroscopy
Year	2020
Preparation	Centrifugal Filtration: The human blood serum samples (n=25) were subjected to centrifugal filtration using Amicon Ultra-0.5 mL centrifugal filter devices with a specific molecular weight cutoff.
	** the samples were collected during routine blood check-ups, 1 mL of the vial remains being provided for further spectroscopic analysis.
Instrument	1. Horiba Jobin-Yvon LabRAM HR800 spectrometer
	2. Olympus IX71 inverted microscope
	3. X60 water immersion objective (LUMPlanF1, Olympus)
	4. Two different laser sources were used: a 532 nm laser and a 785 nm laser
	* In all experiments, a 300 lines/mm grating was used
	* The confocal hole was set at 100 mm for all measurements, the specified setting for confocal operation.
Method	Vibrational spectroscopy
	1. ATR-FTIR
	2. Raman Spectroscopy
Preprocessing	1. Smoothing: the Savitzky–Golay method
	2. Baseline Correction: The rubberband method
	3. Remove spectral interferents: EMSC
	4. Reference Spectra: were used for EMSC and as a basis for quantification.
Results	The paper provides evidence of the potential of Raman spectroscopy for the quantitative analysis of human blood serum constituents.
	Comparison of the results of ATR-FTIR and Raman spectroscopic of patient sample set for monitoring glucose levels are divided as below,
	ATR-FTIR $R^2$ = 0.9957 in concentrartion range 61.25-210 mg/dL
	Raman Spectroscopy $R^2$ = 0.91 in concentrartion range 52.25-210 mg/dL
	Raman Spectroscopy $R^2$ = 0.84 in concentrartion range 52.25-440 mg/dL

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

Title	High-precision prediction of blood glucose concentration utilizing Fourier transform Raman spectroscopy and an ensemble machine learning algorithm
Year	2023
Preparation	1. The blood glucose concentrations of the samples ranged from 4.12 mmol/L - 16.32 mmol/L from hospital
	2. 106 blood samples are separated into 2 groups, the first for the clinical standard testing of blood glucose concentration (32 samples)
	The second group of samples was scanned using Raman spectroscopy to obtain the whole blood Raman spectrum (32 samples)
	** the rests are for training on model (42 samples)
Instrument	1. Raman spectrometer
	The output power of 1064 nm excitation radiation is 90 mW
	The spectrum range is from 400 $cm^{−1}$ to 4000 $cm^{−1}$
	The output power of 1064 nm excitation radiation is 90 mW in the experiment.
	Each sample is scanned at a spectral resolution of 6 $cm^{-1}$
Method	1. Bagging-ABC-ELM prediction model
	2. PLSR (Partial Least Squares Regression) and SVR (Support Vector Regression) models were constructed as traditional methods for comparison.
	** They performed three measuements on every sample for improving accuracy
Preprocessing	1. Baseline correction: Savitzky-Golay denoising
	3. PCA: employed to extract key information from the spectral data and simplify the model input.
	** A total of 24 principal components were selected to represent the original spectral data.
Results	The mean values of $R^2$ and RMSEP for the proposed model were 0.9928 and 0.1928

7. Visible micro-Raman spectroscopy for determining glucose content in beverage industry

Title	Visible micro-Raman spectroscopy for determining glucose content in beverage industry
Year	2011
Preparation	1. Biochemical assay is used to validate the concentrations
	The assay involves the use of glucose oxidase (GOD) from Aspergillus niger to catalyze the oxidation of glucose to gluconic acid, which generates hydrogen peroxide.
	The resulting hydrogen peroxide is detected using a chromogenic oxygen acceptor composed of phenol and 4-aminophenazone in the presence of horseradish peroxidase.
	2. Samples : Sport drinks A,B,C,D
Instrument	Micro-Raman Spectroscopy
	The micro-Raman spectrometer was equipped with an optical confocal microscope (Olympus BX40) connected by a 50 lm optical fiber to a Jobin-Yvon TriAx 180 monochromator equipped
	The visible laser source was a He– Ne laser operating at a wavelength 633 nm, with a maximum nominal power of 17 mW
	50X optical objective (Olympus MPLAN 50X/ 0.75) on a circular area with diameter of 20 lm
	Three gratings with 300, 600 and 1800 grooves/mm were selectable
	The spectra were acquired using accumulation times ranging from 60 to 600 s by means of a double acquisition process which permits the rejection of spurious peaks due to direct CCD excitations
Method	iPLS (interval Partial Least Square)
Preprocessing	--
Results	1. Glucose solution for building iPLS: linear fitting with an angular coefficient equal to 1.004 ± 0.008 and a correlation coefficient of 0.991