Artificial intelligence-driven whole-brain cell mapping with highly multiplexed in situ hybridization.

Area of Science:

• Neuroscience
• Molecular Biology
• Bioimaging

Background:

• Three-dimensional (3D) single-cell-resolution imaging bridges organ-wide and cellular research in development and disease.
• Current whole-organ imaging is limited by restricted molecular marker staining and lack of precise cell population quantification.
• Accurate spatial mapping of cellular ecosystems is crucial for understanding complex biological processes and disease mechanisms.

Purpose of the Study:

• To present a highly multiplexed whole-mount staining technique for 3D tissue imaging.
• To develop an artificial intelligence (AI)-driven workflow for accurate spatial cell mapping.
• To provide a systematic framework for analyzing cellular ecosystems and investigating disease vulnerabilities.

Main Methods:

• Developed mFISH3D (multiplexed Fluorescence In Situ Hybridization in 3D), a technique for repeated application of FISH.
• Applied mFISH3D to visualize 10 types of mRNAs in intact mouse and human brains.
• Integrated an AI-driven workflow for automated spatial cell mapping and quantification, reducing manual annotation.

Main Results:

• mFISH3D enables visualization of multiple mRNA targets in intact organs at single-cell resolution.
• The AI workflow significantly enhances accuracy in spatial cell mapping across large tissue volumes.
• Demonstrated successful application in both mouse and human brain specimens.

Conclusions:

• The integration of mFISH3D and AI provides a powerful framework for analyzing complex cellular ecosystems.
• This approach facilitates comprehensive investigation of cellular vulnerabilities in disease.
• Advances in 3D imaging and AI analysis open new avenues for developmental and disease research.