SKM 2023 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

O: Fachverband Oberflächenphysik

O 12: Scanning Probe Techniques: Method Development I

O 12.4: Vortrag

Montag, 27. März 2023, 15:45–16:00, GER 37

Analysis of DNA-origami-based calibration standard for AFM using artificial intelligence — •Ziba Akbarian^1,2, Tim Johannes Seifert¹, Birka Lalkens², Ingo Busch³, Harald Bosse³, and Uta Schlickum^1,2 — ¹Technische Universität Braunschweig, Braunschweig, Germany — ²Laboratory for Emerging Nanometrology (LENA), Braunschweig, Germany — ³Physikalisch Technische Bundesanstalt, Braunschweig, Germany

In this investigation, we focus on a nanoscale-standard reference system for calibrating atomic force microscopes utilizing DNA origami. To introduce measurable protrusions on the DNA origami nanostructure, markers at well-defined positions have been attached. To analyze the nanostructures as accurately as possible, we propose the use of artificial intelligence (AI) based image analysis techniques. Commonly, evaluation of the obtained data from atomic force microscopes is performed manually and includes preparatory steps to visualize images before analysis. As a result, the data analysis process requires a lot of time and always introduces not-avoidable errors. We present an AI-based measurement procedure using the YOLOv5 object detection and a semantic segmentation network trained on synthetic data to crop origami structures and pixelwise locate markers, respectively. This way, we can achieve a relative calibration accuracy in the range of 1% with an automatic data evaluation tool.