Histopathological H&E-Stained Image Analysis Based on AI (CROSBI ID 718243)
Prilog sa skupa u zborniku | prošireni sažetak izlaganja sa skupa | međunarodna recenzija
Podaci o odgovornosti
Musulin, Jelena ; Štifanić, Daniel ; Zulijani, Ana ; Baressi Šegota, Sandi ; Anđelić, Nikola ; Lorencin, Ivan ; Glučina, Matko ; Car, Zlatan
engleski
Histopathological H&E-Stained Image Analysis Based on AI
Over the past decade, improvements in image analysis methods and substantial advancements in processing power have allowed the development of powerful computer-aided analytical approaches to medical data. Tissue histology slides can now be scanned and preserved in digital form, thanking to the recent introduction of the entire slide digital scanners. In such form, they can serve as input data for Artificial Intelligence (AI) algorithms that can speed up standard procedures for histology analysis with high accuracy and precision. The aim of this research was to create an automated system based on AI for histopathological image analysis. The first step was to normalize H&E-stain images, and then use them as an input to the convolutional neural network. The best results are achieved using ResNet50 with highest AUC value of 0.98 (±σ=0.02). Such an approach proved to be successful in analyzing histopathological images.
artificial intelligence, convolutional neural network, histopathological analysis, oral squamous cell carcinoma
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o prilogu
3
2022.
objavljeno
Podaci o matičnoj publikaciji
First Serbian Conference on Applied Artificial Intelligence
Kragujevac:
Podaci o skupu
1st Serbian International Conference on Applied Artificial Intelligence (SICAAI)
predavanje
19.05.2022-20.05.2022
Kragujevac, Srbija
Povezanost rada
Biotehnologija u biomedicini (prirodno područje, biomedicina i zdravstvo, biotehničko područje), Elektrotehnika, Interdisciplinarne tehničke znanosti, Računarstvo, Temeljne tehničke znanosti