Blood identification at the single-cell level based on a combination of laser tweezers Raman spectroscopy and machine learning

Abstract

Laser tweezers Raman spectroscopy (LTRS) combines optical tweezers technology and Raman spectroscopy to obtain biomolecular compositional information from a single cell without invasion or destruction, so it can be used to “fingerprint” substances to characterize numerous types of biological cell samples. In the current study, LTRS was combined with two machine learning algorithms, principal component analysis (PCA)-linear discriminant analysis (LDA) and random forest, to achieve high-precision multi-species blood classification at the single-cell level. The accuracies of the two classification models were 96.60% and 96.84%, respectively. Meanwhile, compared with PCA-LDA and other classification algorithms, the random forest algorithm is proved to have significant advantages, which can directly explain the importance of spectral features at the molecular level.

Identification of individual red blood cells by single-cell Raman spectroscopy combined with machine learning

“…LTRS system for single-cell laser optical tweezers has been established (Figure 1) [14] . The Olympus optical microscopic objective (100×; N.A.=1.40) can focus a 785 nm near-infrared laser on an optical tweezer of about 1 m to capture a single red blood cell (RBC) and collect the Raman signal of the RBC.…”

“…Fujihara et al used principal component analysis to distinguish blood from 11 species [13] . In our previous study, the random forest model was used to classify the blood of ten species by measuring the Raman spectra of individual red blood cells [14] .…”

Identification of multiple myeloma resistant cells using machine learning and laser tweezers Raman spectroscopy

“…Due to the presence of fluorescent substances in the samples, laser irradiation can induce fluorescence background, which may introduce noise interference into the Raman spectroscopy data. Therefore, before establishing the analysis model, we applied a series of preprocessing steps to eliminate this noise interference [27]. Firstly, we employed the Asymmetric Least Squares method for baseline correction to remove the fluorescence background.…”

Recent advances in surface enhanced Raman spectroscopy for bacterial pathogen identifications

“…Therefore, data learning methods based on machine learning algorithms have emerged by using the pre-learning part of the training data to predict unknown samples. For example, the PCA-LDA algorithm is widely used in Raman spectroscopy data analysis [96] , [97] , [98] . Khulla Naseer et al used PCA algorithm to reduce the dimension of the Raman spectrum data of serum samples infected with Hepatitis C virus (HCV) and healthy people in advance [88] .…”