# Posthandling Spectral Information Enhancement for Single Cell Raman Molecular Mapping Analysis

## TL;DR

This paper introduces a new method to improve Raman molecular mapping of single cells by reducing noise and background interference, enabling clearer identification of minor biochemical features.

## Contribution

A novel objective data analysis workflow for Raman spectroscopy that enhances minor peak detection and mapping accuracy in single-cell studies.

## Key findings

- The proposed workflow enables precise identification of minor Raman peaks in single-cell molecular maps.
- Automated background subtraction improves the accuracy of biochemical component extraction in complex cell environments.
- The method allows quantitative analysis of cellular components without subjective interpretation.

## Abstract

Biochemical analysis of living systems such as single
cells benefits
greatly from the label-free and low-invasive molecular mapping with
Raman microspectroscopy. Sets of Raman spectra at different spatial
points are analyzed to generate Raman molecular maps corresponding
to specific chemical species. However, human error and subjective
data analysis can be technical issues that limit interpretation and
its validity. Here, we present an objective data analysis scheme for
postprocessing large data sets of Raman spectra for molecular mapping
of living cells. The process comprises three steps: (i) Denoising
the spectral data set using low-rank approximation; (ii) obtaining
an objective background from data points outside the target cell;
(iii) subtracting the thus obtained background using Hypothetical
Addition Multivariate Analysis with Numerical Differentiation (HAMAND)
via an automatically determined coefficient. Through the present analysis,
minor Raman peaks, as indiscernible as they are, can be identified
and precisely mapped. We demonstrate a quantitative discussion of
cellular components after extracting contributions only from a single
target cell from a Raman mapping image where multiple cells or parts
of other cells are present. This work opens an improved analysis workflow
for accurate spectroscopic analysis of living cells with the advantage
of identifying minor Raman peaks unambiguously.