Automated Atlas-based Segmentation of Single Coronal Mouse Brain Slices using Linear 2D-2D Registration

Area of Science:

• Neuroscience
• Medical Imaging
• Computational Biology

Background:

• Precise anatomical region identification is vital for brain histological data analysis.
• Manual segmentation is subjective and time-consuming.
• Existing automatic methods often struggle with 2D histological data and 3D atlases.

Purpose of the Study:

• To develop an automated strategy for segmenting 2D coronal histological slices within a 3D digital brain atlas.
• To enable accurate local quantifications and therapeutic evaluations in brain histology.

Main Methods:

• Utilized linear registration for co-registration of 2D histological slices with a 3D digital brain atlas.
• Developed a strategy for automatic 2D-3D segmentation.
• Validated the method's robustness and performance at a whole-brain scale.

Main Results:

• Successfully demonstrated an automated and accurate method for segmenting single 2D coronal slices within a 3D atlas.
• Validated the robustness and performance of the proposed segmentation strategy.
• Provided a solution for analyzing 2D histological data using 3D anatomical references.

Conclusions:

• The proposed automated 2D-3D segmentation strategy effectively addresses the challenge of aligning histological data with digital atlases.
• This method enhances the accuracy and efficiency of brain histological analysis.
• Facilitates improved local quantifications and therapeutic outcome evaluations.