Image-Based AI in the Regulated Pharmaceutical Production

Veranstaltungsdetails
Datum: 06.11.2025, 16:00 Uhr - 18:00 Uhr 
Ort: N 2045, Universitätsstraße 6a, 86159 Augsburg
Veranstalter: Institut für Informatik
Themenbereiche: Studium, Wissenschaftliche Weiterbildung, Informatik, Gesundheit und Medizin
Veranstaltungsreihe: Medical Information Sciences
Veranstaltungsart: Vortragsreihe
Vortragende: Dr. Yvonne Gladbach
BIOINF ASFDASDF DSFASF ASDF ASDF © Universität Augsburg

In diesem Semester wird die im WiSe 2022/23 erfolgreich gestartete Vortragsreihe Medical Information Sciences fortgesetzt. Renommierte Wissenschaftlerinnen und Wissenschaftler unterschiedlicher Fachdisziplinen und Forschungsstandorte geben jeden Donnerstag ab 16:00 Uhr Einblicke in aktuelle Fragestellungen und Anwendungsgebiete des breiten Forschungsfeldes Medical Information Sciences.


Artificial intelligence (AI) is increasingly integrated into regulated industrial environments, such as pharmaceutical manufacturing. At a production site for the synthesis of active pharmaceutical ingredients (APIs), an image-based AI system was developed to support microbiological process control. This system applies random forest algorithms trained on microscopic images to assess the morphology of microorganisms involved in biotechnological processes for in-process contamination detection.

The AI system functions as a decision-support tool, enabling laboratory experts to identify signs of contamination or process instability. Through real-time assessment and pattern recognition, it enhances process robustness, reduces the risk of production loss, and supports continuous quality in accordance with Good Manufacturing Practice (GMP) standards.

Unlike generative AI models, this AI is specifically designed for supervised visual analysis, ensuring full human oversight and compliance with relevant regulations. Different aspects matter, such as the development pipeline, training and validation methodology, and integration of the model into an existing GMP framework, illustrating how image-based AI can improve reliability and efficiency in pharmaceutical production without compromising patient safety.

Weitere Veranstaltungen der Veranstaltungsreihe "Medical Information Sciences"

Weitere Veranstaltungen: Institut für Informatik


Suche