AI-assisted workflow that automated TEM data analysis reduces materials knowledge extraction to minutes

1/ AI usage method: AI-assisted analysis workflow including (S)TEM image segmentation and FFT peak detection
2/ Application task: Automation of data analysis including phase identification, strain map calculation, and 3D model generation from electron microscopy images
3/ Reported result: Analysis of (S)TEM that took experts days was reported to be reduced to minutes
4/ The paper published on 2026-09-11 presented an AI-assisted workflow for (S)TEM data analysis.
5/ This workflow automated the extraction of materials information from electron microscopy data by connecting image segmentation, FFT peak detection, crystal phase identification, strain map calculation, and 3D model generation.
6/ The paper reported that analysis that previously took experts days was reduced to minutes.
7/ From the perspective of materials research, the key point of this result is that complex analyses that are difficult for humans to process one by one, such as high-resolution microscopy data, can be handed over to an AI-based bundle of steps.
8/ In particular, it is practical that different analyses such as microstructure identification, phase identification, and strain distribution calculation are connected into a single flow.
9/ The limitations are also clear.
10/ The performance of such a workflow can be strongly affected by the resolution of the input data, noise, specimen alignment state, and the criterion settings of each step.
11/ Since the time savings reported in the paper cannot be assumed to apply identically to all TEM data, the accuracy and processing time should be rechecked under the sample and instrument conditions being used. [1][2]