Related Experiment Video
Updated: May 1, 2026
14:14
Targeting Neuronal Fiber Tracts for Deep Brain Stimulation Therapy Using Interactive, Patient-Specific Models
Published on: August 12, 2018
9.3K
Seeing the Middle: Reconstructing 3D Internal Electrode Microstructures from Low-Resolution Surfaces with Generative
Zhiqiang Niu1, Zhaoxia Zhou2, Patrice Perrenot3
1Department of Aeronautical and Automotive Engineering Loughborough University Loughborough LE11 3TU UK.
Small Science
|November 19, 2025
Summary
REMind, an AI model, reconstructs 3D energy material microstructures from limited data, overcoming microscopy limitations. This accelerates the characterization of advanced energy materials for improved performance and durability.
Area of Science:
- Materials Science
- Artificial Intelligence
- Energy Technology
Background:
- Characterizing 3D complex energy materials interfaces is crucial for understanding structure-property relationships.
- Current microscopy techniques face limitations in resolution and speed, hindering high-throughput analysis of energy materials.
Purpose of the Study:
- To present REMind, a generative AI model for rapid and accurate reconstruction of electrode microstructures.
- To enable high-fidelity imaging of energy materials with reduced data acquisition requirements.
Main Methods:
- Developed REMind, a generative diffusion AI model trained on high-resolution microstructures.
- Utilized focused ion beam-scanning electron microscopy (FIB-SEM) for data acquisition.
- Employed a multi-scale multi-physics SOFC model to quantify electrochemical performance impact.
Main Results:
- REMind accurately reconstructs internal microstructures with low pixel-wise error (<10%).
- The model quantifies reconstruction uncertainty using generated entropy.
- Validated for solid oxide fuel cell (SOFC) anodes, proton exchange membrane fuel cells, and solid-state batteries.
Conclusions:
- REMind significantly enhances the high-throughput characterization of energy materials.
- The AI model demonstrates broad applicability across various energy technologies.
- REMind facilitates a deeper understanding of energy material performance and degradation.
