This collection of articles highlights the transformative impact of advanced imaging and deep learning technologies on cancer research and clinical oncology. The first article reviews the integration of deep learning with spatial omics technologies, emphasizing its growing role in biomarker discovery, image analysis, and clinical decision-making by enabling more accurate cell segmentation, phenotyping, and prognostication from highly multiplexed imaging data. The second study demonstrates the clinical significance of spatial organization and phenotypes of immune cells, specifically CD8 T cells, within the tumor microenvironment of oropharyngeal cancer patients. Using multiplex immunofluorescence and sophisticated spatial analysis, the study links spatial patterns of immune infiltration to improved patient outcomes, while also noting that tumor stem cell markers require further evaluation. The third article introduces RAPID, a GPU-accelerated software platform that streamlines the processing and analysis of increasingly large and complex multiplexed fluorescence microscopy data, addressing key challenges in image registration, deconvolution, and artifact reduction. Collectively, these works showcase the advances in computational and imaging tools driving deeper biological insight, more precise risk stratification, and enhanced clinical applications in oncology.
----
In the past several years we have seen an explosion in technologies that enable multiplexed characterization of single-cell biology in situ for both RNA and proteins. These techniques have been applied across the fields of microbiology, immunology and oncology. Similarly, international consortia have formed to use this technology to create atlases of healthy and diseased tissues. This has created new spatial biology datasets. These new datasets enable new computational analyses and require new analytical approaches to extract meaning at the single cell level.
This Research Topic focuses on the use of computational approaches to:
1) Process multiplexed imaging data to extract biologically and clinically relevant information of tissues
2) Use multiplexed imaging data for models or training algorithms
3) Develop spatial statistics to describe biological process, and/or
4) Develop methods to characterize cell types, tissue regions, or disease states.
5) Extract cell-cell interactions critical for immunological or pathological projects.
Some examples include:
1) The application of machine learning principles to image processing, cell segmentation, or even identification of cell types.
2) Agent-based or multiscale modeling approaches.
3) Spatial statistics and descriptions of multicellular units
4) Data compression or presentation formats for multiplexed images
5) Extraction and quantification of non-cellular components like the extracellular matrix
6) Integration with other datasets on the same tissue such as single cell RNA sequencing
7) Application of graph-based tools to understand nodes and connections in the system
8) 3D reconstruction from 2D images
9) integration of multiple imaging modalities
10) New batch correction methods, and
11) The role of morphological analysis in cell-type identification.
12) Establishing methods for taking into account the mobility of immune cells in fixed tissues.
Statement: C.M.S. is a scientific advisor to, has stock options in, and has received research funding from Enable Medicine, Inc.
Keywords: Multiplexed Imaging, Machine learning, single-cell, bioinformatics, modeling, immunology
Important note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
