AI-driven projection tomography with multicore fibre-optic cell rotation.

Area of Science:

• Biophysics
• Cell Biology
• Medical Imaging

Background:

• Optical tomography offers 3D insights into subcellular structures but faces limitations in scanning range, causing anisotropic resolution.
• Conventional methods require manual processing, hindering efficiency in cellular imaging and analysis.

Purpose of the Study:

• To develop a novel optical tomography approach for full-angle projection with isotropic resolution.
• To introduce an AI-driven workflow for autonomous tomographic reconstruction, overcoming manual processing limitations.
• To demonstrate the application of this technique for 3D reconstruction of cellular structures.

Main Methods:

• Utilized a compact multi-core fiber-optic cell rotator system for precise optical manipulation within a microfluidic chip.
• Implemented an AI-driven tomographic reconstruction workflow for autonomous data processing.
• Validated the approach using cell phantoms and HL60 human cancer cells.

Main Results:

• Achieved full-angle projection tomography with isotropic resolution, overcoming conventional limitations.
• Demonstrated successful 3D reconstruction of cell phantoms and human cancer cells.
• Showcased the AI workflow's autonomous and efficient performance.

Conclusions:

• The AI-driven cell rotation tomography provides a powerful, non-invasive tool for high-resolution 3D cellular imaging.
• This versatile approach has broad applications in diverse tomographic modalities, advancing cell biology and diagnostics.
• Potential to accelerate therapeutic development and improve early-stage cancer detection.