Publikationsdatenbank

Application of artificial intelligence for quantification of angiogenesis (2022)

Art

Poster

Autoren

Alshamy, Zaher (WE 1)

Plendl, Johanna (WE 1)

Kässmeyer, Sabine

Kongress

116th Annual Meeting of the Anatomische Gesellschaft : joint meeting with the Anatomical Society

Berlin, 20. – 23.09.2022

Quelle

Abstract book : 116th annual meeting of the Anatomische Gesellschaft : joint meeting with the Anatomical Society — Charité - Universitätsmedizin Berlin (Hrsg.)

— S. 5

Sprache

Englisch

Verweise

URL (Volltext): https://anatomical.aey-congresse.de/index/program.html?file=files/anatomical/202209_Berlin_116th-Meeting_AbstractBook.pdf&cid=10390

Kontakt

Institut für Veterinär-Anatomie

Koserstr. 20
14195 Berlin
+49 30 838 75784
anatomie@vetmed.fu-berlin.de

Abstract / Zusammenfassung

Angiogenesis is the process of new blood vessel formation. Manual methods, which are most commonly used to evaluate angiogenesis in vitro, are time-consuming and demand considerable effort. This study aims to implement a new artificial intelligence method for quantifying angiogenesis reliably and fast.

Human skin endothelial cells were cultured in nutrient medium and labelled with the endothelial marker anti-CD31. The cells proliferated and formed a 3D tubular network.
Number and diameter of the capillary-like structures (endothelial tubes), as well as their crossing points (knots) were manually counted as control.
Applying the AI-module (Segment.ai), a small set of hand-traced microscopic images of angiogenesis was used to train the software in the first phase (train phase). Training results were applied on similar images in the second phase (application phase) to identify various parameters.

Number and diameter of endothelial tubes were similar in both methods. For the manual method, mean of numbers and diameter of endothelial tubes were 835.17 ± 52.37 SEM and 9.91 μm ± 0.21 SEM, respectively; for the AI method 865 ± 103.58 SEM and 10.02 μm ± 0.80 SEM, respectively. Number of knots was 323 ± 26 for manual and 610 ± 159 for AI. Training the AI took about 50 hours. Time used for quantification per image was about 40 minutes for the manual and 5 minutes for AI method.

The artificial intelligence software is a suitable method for quantifying angiogenesis in vitro. Compared to the manual method, it saves time and is more efficient for experienced users.