Comprehensive analysis of human dendritic spine morphology and density

Area of Science:

• Neuroscience
• Cell Biology
• Computational Biology

Background:

• Dendritic spines are crucial for brain function, modulating neural activity through morphological changes.
• In-depth analysis of dendritic spines in human brain tissue has been limited.
• Understanding human dendritic spine morphology is vital for neurological and psychiatric disease research.

Purpose of the Study:

• To comprehensively analyze human dendritic spine morphology and density using a unique human brain tissue dataset.
• To evaluate a deep learning model for automated 3-D spine segmentation and reconstruction.
• To investigate sex- and tissue-specific differences in dendritic spine characteristics.

Main Methods:

• Utilized acute slices and organotypic brain slice cultures from 27 human patients.
• Performed 3-D reconstruction of dendritic spines using ZEISS arivis Pro software.
• Developed and applied a deep learning model for automated spine segmentation and 3-D reconstruction.

Main Results:

• Identified significant differences in spine density by sex (females > males), dendrite type (apical > basal), and tissue condition (acute > cultured).
• Observed changes in spine morphology over time in culture: mushroom spines decreased, while stubby and thin spines increased.
• The deep learning model achieved a 74% F1-score and reduced processing time by over 50%.

Conclusions:

• Human brain tissue analysis reveals unique synaptic properties and sex/tissue-specific dendritic spine dynamics.
• Integrating deep learning with traditional methods enables efficient, large-scale analysis of dendritic spines.
• Findings provide insights into potential mechanisms of neurological and psychiatric diseases related to spine morphology.