Serum Raman Spectroscopy Data for COVID-19 Screening from 157 Patients

Gang Yin provides a dataset of 2,355 Raman spectra collected from 157 serum samples. The samples were obtained from 53 confirmed COVID-19 patients, 54 suspected cases, and 50 healthy controls. The data was used to develop a diagnostic algorithm using support vector machine (SVM) methods.

Use Cases

• Developing classification models for COVID-19 based on Raman spectral features from serum.
• Comparing spectral signatures between confirmed patients, suspected cases, and healthy controls.
• Reproducing the analysis from the related manuscript using the provided raw data and MATLAB code.
• Exploring the utility of Raman spectroscopy as a primary screening tool for viral infections.

Strengths

• Includes raw spectral data from 157 individuals, totaling 2,355 spectra.
• Provides separate data files for confirmed COVID-19 patients (53), suspected cases (54), and healthy controls (50).
• Dataset is accompanied by the original MATLAB processing code (code.m).

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• The primary data file (data.mat) is in a proprietary .mat format requiring MATLAB software.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

Gang Yin via paperswithcode

Collection Method

Raman spectroscopy analysis of serum samples using a Volume Phase Holographic (VPH) spectrograph and deep-cooled CCD camera.