Abstract:Surface‐enhanced Raman scattering nanoparticles (SERS NPs) are powerful imaging contrast agents. Multiplexed spectral images are commonly unmixed with reference spectra to calculate quantitative image channels. Controlled images of SERS NP mixtures are evaluated with a variety of unmixing algorithms to reveal differences in their quantitation properties. Surface‐enhanced Raman scattering nanoparticles (SERS NPs) offer powerful optical contrast features for imaging assays. Their gold core enhances the inelastic scattering cross section, allowing highly sensitive and rapid detection, and their characteristic sets of narrow spectral bands give them unsurpassed multiplexing capabilities. Multiplexed hyperspectral images are commonly unmixed using a compensation matrix of reference spectra to produce quantitative image channels illustrating the distribution of each material. It is these unmixed channels that are fit for interpretation from assays utilizing SERS NP contrast agents. Some factors that may impact SERS NP quantitative and dynamic range capabilities may include endogenous background heterogeneity, the ability of unmixing algorithms to account for signal variances, and linear system conditioning imposed by contrast agent signals. We report on hyperspectral Raman imaging of mixtures of SERS NPs from an expanded library of contrast agents. We study increasing plexity and varying degrees of system conditioning as inputs to a diverse set of classical, non‐negatively constrained, and regularized regression algorithms to investigate which signal features and unmixing methods deliver the most promising quantitation performance with the least error. Raman imaging of SERS NP mixtures is performed on controlled substrates and representative biological specimens, and experimental results are compared against ground truth data. We evaluate spectral fitting fidelity, quantitation, and specificity correlations with system conditioning. Spectral unmixing with a regularized hybrid of least squares regression with principal component analysis (HLP) algorithm approximated spectra with 3.5× better fitting fidelity and 3× better quantitation robustness with tissue background compared with simpler unmixing routines.

Leveraging Machine Learning for Advanced Nanoscale X-ray Analysis: Unmixing Multicomponent Signals and Enhancing Chemical Quantification

From STEM-EDXS data to phase separation and quantification using physics-guided NMF

Machine Learning Approach to Enable Spectral Imaging Analysis for Particularly Complex Nanomaterial Systems

Highly sensitive 2D X-ray absorption spectroscopy via physics informed machine learning

Non-negative matrix factorization-aided phase unmixing and trace element quantification of STEM-EDXS data

Retrieving the quantitative chemical information at nanoscale from SEM EDX measurements by Machine Learning

Deep Learning for Spectroscopic X‐ray Nano‐Imaging Denoising

Nanoscale chemical imaging with structured X-ray illumination

NNLSF: A Fast and Informative Fitting Method for XANES Chemical Mapping Analysis.

Robust retrieval of material chemical states in X-ray microspectroscopy

A probabilistic deep learning approach to automate the interpretation of multi-phase diffraction spectra

Subspace modeling for fast and high-sensitivity X-ray chemical imaging

Imaging Atomic-Scale Chemistry from Fused Multi-Modal Electron Microscopy

Towards scanning nanostructure X‐ray microscopy

Assessment of unmixing approaches for the quantitation of SERS nanoparticles in highly multiplexed spectral images

Predicting and Accelerating Nanomaterial Synthesis Using Machine Learning Featurization

Creating 3D Nanoparticle Structural Space via Data Augmentation to Bidirectionally Predict Nanoparticle Mixture's Purity, Size, and Shape from Extinction Spectra

Enhanced Nanoparticle Recognition via Deep Learning-Accelerated Plasmonic Sensing

A machine learning model for textured X-ray scattering and diffraction image denoising

Multi-modal Machine Learning Analysis of X-ray Absorption Near-Edge Spectra and Pair Distribution Functions: Performance and Interpretability towards Experimental Design