Detection of NS1 from SERS spectra using K-NN integrated with PCA

Othman, N. H.; Lee, Khuan Y.; Radzol, A. R. M.; Mansor, W.; Rashid, Umer

doi:10.1109/iecbes.2016.7843421

Cited by 2 publications

(3 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…38−40 For example, the k-nearest neighbor (k-NN) is appropriate for a small number of parameters on uniformed cells or preselected cell groups with single or multiple biophysical sensing techniques. 36,41 The simple principal component analysis (PCA) or linear discriminant analysis (LDA) has been applied to specific types of data analysis, such as Raman spectroscopy on tumor cells. 42−45 The support vector machine (SVM), Lasso, ENet, or nonnegative garrote on kernel machine (NGK) are suitable for multiple dimensional biophysical properties analysis, such as biomechanical, bioelectrical, biochemical, and bio-optical sensing data.…”

Section: ■ Resultsmentioning

confidence: 99%

“…There is no single machine learning-based algorithm that is suitable for all types of biophysical properties data. Most of the current machine learning algorithms are applied to cells or biomarker imaging in glioma diagnosis. − For example, the k-nearest neighbor (k-NN) is appropriate for a small number of parameters on uniformed cells or preselected cell groups with single or multiple biophysical sensing techniques. , The simple principal component analysis (PCA) or linear discriminant analysis (LDA) has been applied to specific types of data analysis, such as Raman spectroscopy on tumor cells. − The support vector machine (SVM), Lasso, ENet, or nonnegative garrote on kernel machine (NGK) are suitable for multiple dimensional biophysical properties analysis, such as biomechanical, bioelectrical, biochemical, and bio-optical sensing data. ,,, Here, we demonstrated the ENet and Lasso methods on biomechanical data of different grades of glioma cells with over 80% prediction accuracy. Lasso and ENet have been developed when p ≪ n cases and used in various cancer studies. , Although our sample size is relatively small for biological findings, the number of variables is still smaller than the sample size, that is, p < n in our study.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Glioma Single-Cell Biomechanical Analysis by Cyclic Conical Constricted Microfluidics

Geng,

Zhou,

et al. 2023

Anal. Chem.

View full text Add to dashboard Cite

Determining the grade of glioma is a critical step in choosing patients' treatment plans in clinical practices. The pathological diagnosis of patient's glioma samples requires extensive staining and imaging procedures, which are expensive and timeconsuming. Current advanced uniform-width-constriction-channelbased microfluidics have proven to be effective in distinguishing cancer cells from normal tissues, such as breast cancer, ovarian cancer, prostate cancer, etc. However, the uniform-width-constriction channels can result in low yields on glioma cells with irregular morphologies and high heterogeneity. In this research, we presented an innovative cyclic conical constricted (CCC) microfluidic device to better differentiate glioma cells from normal glial cells. Compared with the widely used uniform-width-constriction microchannels, the new CCC configuration forces single cells to deform gradually and obtains the biophysical attributes from each deformation. The human-derived glioma cell lines U-87 and U-251, as well as the human-derived normal glial astrocyte cell line HA-1800 were selected as the proof of concept. The results showed that CCC channels can effectively obtain the biomechanical characteristics of different 12−25 μm glial cell lines. The patient glioma samples with WHO grades II, III, and IV were tested by CCC channels and compared between Elastic Net (ENet) and Lasso analysis. The results demonstrated that CCC channels and the ENet can successfully select critical biomechanical parameters to differentiate the grades of single-glioma cells. This CCC device can be potentially further applied to the extensive family of brain tumors at the single-cell level.

show abstract

Section: ■ Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Glioma Single-Cell Biomechanical Analysis by Cyclic Conical Constricted Microfluidics

Geng,

Zhou,

et al. 2023

Anal. Chem.

View full text Add to dashboard Cite

show abstract

“…Since the Raman peaks at different wavenumbers can be viewed as high-dimensional variables, machine learning algorithms have been applied to analyze the data from SERS spectra. J. Liu et al published studies on cancer research using SERS nanoparticles for Raman imaging and predicted that machine learning methods can accurately identify tumor SERS imaging, which represents unique biomarker expression signatures at the molecular level 20 21 . However, the k-NN algorithm is nonparametric without a particular model to fit.…”

Section: Introductionmentioning

confidence: 99%

Scalable nanolaminated SERS multiwell cell culture assay

et al. 2020

View full text Add to dashboard Cite

This paper presents a new cell culture platform enabling label-free surface-enhanced Raman spectroscopy (SERS) analysis of biological samples. The platform integrates a multilayered metal-insulator-metal nanolaminated SERS substrate and polydimethylsiloxane (PDMS) multiwells for the simultaneous analysis of cultured cells. Multiple cell lines, including breast normal and cancer cells and prostate cancer cells, were used to validate the applicability of this unique platform. The cell lines were cultured in different wells. The Raman spectra of over 100 cells from each cell line were collected and analyzed after 12 h of introducing the cells to the assay. The unique Raman spectra of each cell line yielded biomarkers for identifying cancerous and normal cells. A kernel-based machine learning algorithm was used to extract the high-dimensional variables from the Raman spectra. Specifically, the nonnegative garrote on a kernel machine classifier is a hybrid approach with a mixed nonparametric model that considers the nonlinear relationships between the higher-dimension variables. The breast cancer cell lines and normal breast epithelial cells were distinguished with an accuracy close to 90%. The prediction rate between breast cancer cells and prostate cancer cells reached 94%. Four blind test groups were used to evaluate the prediction power of the SERS spectra. The peak intensities at the selected Raman shifts of the testing groups were selected and compared with the training groups used in the machine learning algorithm. The blind testing groups were correctly predicted 100% of the time, demonstrating the applicability of the multiwell SERS array for analyzing cell populations for cancer research.

show abstract

Detection of NS1 from SERS spectra using K-NN integrated with PCA

Cited by 2 publications

References 30 publications

Glioma Single-Cell Biomechanical Analysis by Cyclic Conical Constricted Microfluidics

Glioma Single-Cell Biomechanical Analysis by Cyclic Conical Constricted Microfluidics

Scalable nanolaminated SERS multiwell cell culture assay

Contact Info

Product

Resources

About