Understanding mass spectrometry images: complexity to clarity with machine learning

Abstract

The application of artificial intelligence and machine learning to hyperspectral mass spectrometry imaging (MSI) data has received considerable attention over recent years. Various methodologies have shown great promise in their ability to handle the complexity and size of MSI data sets. Advances in this area have been particularly appealing for MSI of biological samples, which typically produce highly complicated data with often subtle relationships between features. There are many different machine learning approaches that have been applied to MSI data over the past two decades. In this review, we focus on a subset of non-linear machine learning techniques that have mostly only been applied in the past 5 years. Specifically, we review the use of the self-organizing map (SOM), SOM with relational perspective mapping (SOM-RPM), t-distributed stochastic neighbor embedding (t-SNE) and uniform manifold approximation and projection (UMAP). While not their only functionality, we have grouped these techniques based on their ability to produce what we refer to as similarity maps. Similarity maps are color representations of hyperspectral data, in which spectral similarity between pixels-that is, their distance in high-dimensional space-is represented by relative color similarity. In discussing these techniques, we describe, briefly, their associated algorithms and functionalities, and also outline applications in MSI research with a strong focus on biological sample types. The aim of this review is therefore to introduce this relatively recent paradigm for visualizing and exploring hyperspectral MSI, while also providing a comparison between each technique discussed.

Spectral organ fingerprints for intraoperative tissue classification with hyperspectral imaging

“…t-distributed Stochastic Neighbor Embedding (t-SNE) [10] is a machine learning method commonly used to reduce the number of dimensions of high-dimensional data and was used to visualize the characteristic reflectance spectra of each pig organ. This non-linear multi-dimensionality reduction tool has already proven valuable for the analysis of HSI and mass spectrometry data [61] and was chosen for visualization as it has shown particular promise for biological samples in the past [62, 63]. The algorithm aims at modelling manifolds of high-dimensional data, and produces low-dimensional embeddings that are optimized for preserving the local neighbourhood structure of the high-dimensional manifold [10].…”

Multi-Class Cancer Subtyping in Salivary Gland Carcinomas with MALDI Imaging and Deep Learning

“…Using the predicted masses as features for a nonlinear dimensionality reduction method (NLDR) provides an extra layer of interpretability to the otherwise opaque classification approach. NLDR methods such as UMAP are a useful tool to visualize high-dimensional data [ 30 ]. We applied densMAP, an augmented version of UMAP which has shown to provide better clustering with regard to the local density of the samples in the original space.…”

Identification of Immune-Related Genes Associated With Bladder Cancer Based on Immunological Characteristics and Their Correlation With the Prognosis

“…Then, we used the hierarchical clustering method of Euclidean distance and Ward linkage to do the immune stratification of BLCA patients. Meanwhile, we also made use of the T-distribution stochastic neighbor embedding (tSNE) algorithm to determine the immune stratification of BLCA patients through RtSEN package (Gardner et al, 2021). (C) Evaluation of tumor immune microenvironment: based on ESTIMATE algorithm, BLCA transcriptome data was utilized to predict stromal cell score, immune cell score and tumor purity, and then the content of these two types of cells was predicted, from which StromalScore, ImmuneScore and EstimateScore were determined (Yoshihara et al, 2013).…”