AI Image Analytics

AI Image Analytics

Artificial-Intelligence Image Analytics

DigiM team has extensive experience in image processing. We help our clients to solve the toughest image processing problems efficiently and reproducibly. Whenever needed, we  push the boundary of the state-of-the-art software algorithms, 

Image segmentation transforms greyscale values from images to material phases. Without accurate segmentation, further analysis will be in a garbage-in-garbage-out situation. DigiM team uses an iterative AI approach to image segmentation, as illustrated in the table below. Quantifications can be conducted on segmented images, as well as greyscale images. Quantitative assessment includes, 

  • Geometric quantification of microstructures (pores, defects, inclusions, particles, fibers, layers) in terms of their volume fraction, size distribution, connectivity, orientation, surface area, and more
  • Evaluation of microstructure design in correlation with product performance, engineering process, and manufacturing conditions
  • Quality control of microstructure (representativeness, uniformity, and composition)

DigiM AI analytical workflow is in compliance FDA 21 CFR Part 11. The results are reproducible with full audit trail support. All intermediate results, including segmentation, can be reviewed by the user. 

DigiM AI image analytics detecting recrystallization in an amorphous solid dispersion
  1. Tracey, J., Lin, S., Jankovic, J., Zhu, A., Zhang, S. (2019). Iterative Machine Learning Method for Pore-Back Artifact Mitigation in High Porosity Membrane FIB-SEM Image Segmentation. Microscopy and Microanalysis
  2. Zhang, S., Byrnes, A. P., Jankovic, J., & Neilly, J. (2019). Management, Analysis, and Simulation of Micrographs with Cloud Computing. Microscopy Today, 27(2), 26-33.
  3. Cheney C.Y. Zhang. Reconstruction of Three-Dimensional Micro-Structures From Two-Dimensional Microscopic Images Using Texture Synthesis and Phase Field Method. Poster ID PDP-54. 
  4. DigiM Artificial Intelligence Image Processing. DigiM Technology Highlight 2017 July Issue. July 29, 2017. 
  5. DigiM FIB-SEM Curtaining Removal. DigiM Technology Highlight. February 25, 2015.
  6. Zhang, S. (2014). Integrated Material Characterization and Property Prediction Using 3D Image-based Analytics and Modeling. Presented at TMS 2014 Annual Meeting in San Diego, CA (February 16-20, 2014).