Related Experiment Video
Updated: Oct 5, 2025
Whole-Brain Single-Cell Imaging and Analysis of Intact Neonatal Mouse Brains Using MRI, Tissue Clearing, and Light-Sheet Microscopy
Published on: August 1, 2022
MIRIAM: A machine and deep learning single-cell segmentation and quantification pipeline for multi-dimensional tissue
Eliot T McKinley1,2, Justin Shao1,2, Samuel T Ellis1
1Epithelial Biology Center, Vanderbilt University Medical Center, Nashville, Tennessee, USA.
Abstract:
Increasingly, highly multiplexed tissue imaging methods are used to profile protein expression at the single-cell level. However, a critical limitation is the lack of robust cell segmentation tools for tissue sections. We present Multiplexed Image Resegmentation of Internal Aberrant Membranes (MIRIAM) that combines (a) a pipeline for cell segmentation and quantification that incorporates machine learning-based pixel classification to define cellular compartments, (b) a novel method for extending incomplete cell membranes, and (c) a deep learning-based cell shape descriptor. Using human colonic adenomas as an example, we show that MIRIAM is superior to widely utilized segmentation methods and provides a pipeline that is broadly applicable to different imaging platforms and tissue types.
Insights
We developed Multiplexed Image Resegmentation of Internal Aberrant Membranes (MIRIAM), a new tool for cell segmentation in tissue imaging. MIRIAM improves accuracy for single-cell protein profiling in complex biological samples.
Area of Science:
- Computational biology
- Biomedical imaging
- Pathology
Background:
- Multiplexed tissue imaging enables single-cell protein profiling.
- Accurate cell segmentation is crucial but challenging for tissue sections.
- Existing segmentation tools often lack robustness.
Purpose of the Study:
- To introduce Multiplexed Image Resegmentation of Internal Aberrant Membranes (MIRIAM), a novel cell segmentation pipeline.
- To address limitations in current cell segmentation methods for multiplexed tissue imaging.
- To provide a broadly applicable solution for tissue section analysis.
Main Methods:
- Developed a machine learning-based pixel classification pipeline for cellular compartment definition.
- Implemented a novel method for extending incomplete cell membranes.
- Integrated a deep learning-based cell shape descriptor.
- Applied the pipeline to human colonic adenomas.
Main Results:
- MIRIAM demonstrated superior performance compared to widely used segmentation methods.
- The pipeline effectively segments cells and quantifies protein expression at the single-cell level.
- Validated the method's applicability across different imaging platforms and tissue types.
Conclusions:
- MIRIAM offers a robust and accurate solution for cell segmentation in multiplexed tissue imaging.
- This tool enhances the analysis of protein expression in complex tissue architectures.
- MIRIAM is a versatile pipeline applicable to diverse research and clinical settings.
