Chair for Materials Science and Engineering
Institute for Materials • Faculty of Mechanical Engineering
Project AutoMorph
Automated Morphological Characterization of Pharmaceutical Excipients and Active Ingredients Using AI-Powered Image Data Analytics
In the AutoMorph project, the use of computer vision technology combined with artificial intelligence (AI) aims to enable the automated evaluation of micrographs for the precise determination of particle properties. The project includes semantic segmentation, detection, classification, and geometric analysis of microstructural objects. To this end, a database will be created that contains microscopy images of pharmaceutical excipients and active ingredients along with associated target parameters such as particle properties or segmentation, which will serve as the basis for training AI models. The AI-based particle size determination, which is already used for the microstructural characterization of metallic materials, is now set to be applied in the field of pharmacy. The goal of AutoMorph is the automated morphological characterization of semi-solid and powder formulations, promising the chemical, especially the pharmaceutical industry, reduced effort in quality assurance and more precise analyses. AI modules can be tailored for various applications and characterization parameters. The standardization of data analysis across imaging techniques and application areas through an expandable ontology represents an innovation in the industrial application of AI.
