Comparison of segmentation techniques for histopathological images - Université de Lorraine
Communication Dans Un Congrès Année : 2015

Comparison of segmentation techniques for histopathological images

Résumé

Image segmentation is a widely used in medical imaging applications by detecting anatomical structures and regions of interest. This paper concerns a survey of numerous segmentation model used in biomedical field. We organized segmentation techniques by four approaches, namely, thresholding, edge-based, region-based and snake. These techniques have been compared with simulation results and demonstrated the feasibility of medical image segmentation. Snake was demonstrated a capability with a high performance metrics to detect irregular shape as carcinoma cell type. This study showed the advantage of the deformable segmentation technique to segment abnormal cells with Dice similarity value over 83%.

Domaines

Informatique
Fichier non déposé

Dates et versions

hal-01785607 , version 1 (04-05-2018)

Identifiants

Citer

Hawraa Haj-Hassan, Ahmad Chaddad, Camel Tanougast, Youssef Harkouss. Comparison of segmentation techniques for histopathological images. 2015 Fifth International Conference on Digital Information and Communication Technology and its Applications (DICTAP), Apr 2015, Beirut, Lebanon. ⟨10.1109/DICTAP.2015.7113175⟩. ⟨hal-01785607⟩
22 Consultations
0 Téléchargements

Altmetric

Partager

More