Computational delineation and cellular profiling of murine cortical cell layers using multiplex immunofluorescence...

Area of Science:

• Neuroscience
• Computational Biology
• Biotechnology

Background:

• The mammalian cerebral cortex exhibits a six-layered vertical organization, with each layer performing specific functions.
• Accurate identification of cortical neurons within their respective layers is crucial for understanding brain function and disease.

Purpose of the Study:

• To present a novel data-driven method for delineating cortical cell layers in brain sections.
• To enable unbiased, quantitative profiling of cortical layers based on cellular composition, phenotype, and spatial arrangement.

Main Methods:

• Utilized an Actively Informed Dirichlet Process Mixture Model (AIDPMM), a machine learning approach, for spatial cluster analysis of neuronal features.
• Applied the method to coronal brain sections imaged with multiplexed immunofluorescence microscopy.
• Enabled parcellation of cytometric measurements by cortical layer for comprehensive analysis.

Main Results:

• Achieved high accuracy in cortical layer delineation, with 92.5% intersection over union (IoU) for layer-specific markers and an R2 of 93.5% concordance with human delineations.
• Demonstrated layer-specific microglia and astrocyte activation in a traumatic brain injury model.
• Showed that lithium+valproate treatment modified injury-induced layer-specific cellular responses.

Conclusions:

• The AIDPMM method is efficient, versatile, and suitable for visual inspection and proofreading.
• Provides an unbiased approach for delineating cortical neuronal layers and profiling cytometric data.
• Facilitates cell layer-specific analysis of experimental manipulations and disease models.