# Neural Network-Based Filter Design for Compressive Raman Classification of Cells

## TL;DR

This paper introduces a neural network method to speed up Raman spectroscopy for cell classification, making it practical for high-throughput cell-based therapies.

## Contribution

A novel neural network approach for optimizing compressive Raman sensing parameters to enable fast, accurate cell classification.

## Key findings

- The method achieves 90% classification accuracy using only five Raman intensity combinations.
- It reduces Raman measurement time by two orders of magnitude.
- The approach is demonstrated on three different cell types.

## Abstract

Cell-based therapies are bound to revolutionize medicine,
but significant
technical hurdles must be overcome before wider adoption. In particular,
nondestructive, label-free methods to characterize cells in real time
are needed to optimize the production process and improve quality
control. Raman spectroscopy, which provides a fingerprint of a cell’s
chemical composition, would be an ideal modality but is too slow for
high-throughput applications. Compressive Raman techniques, which
measure only linear combinations of Raman intensities, can be fast
but require careful optimization to deliver high performance. Here,
we develop a neural network model to identify optimal parameters for
a compressive sensing scheme that reduces measurement time by 2 orders
of magnitude. In a data set containing Raman spectra of three different
cell types, it achieves up to 90% classification accuracy using only
five linear combinations of Raman intensities. Our method thus unlocks
the power of Raman spectroscopy for the characterization of cell products.