SCOPE: Structural Continuity Preservation for Retinal Vessel Segmentation

Yousef Yeganeh, Göktuğ Güvercin, Rui Xiao, Amr Abuzer, Ehsan Adeli, Azade Farshad, Nassir Navab

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

Abstract

Although the preservation of shape continuity and physiological anatomy is a natural assumption in the segmentation of medical images, it is often neglected by deep learning methods that mostly aim for the statistical modeling of input data as pixels rather than interconnected structures. In biological structures, however, organs are not separate entities; for example, in reality, a severed vessel is an indication of an underlying problem, but traditional segmentation models are not designed to strictly enforce the continuity of anatomy, potentially leading to inaccurate medical diagnoses. To address this issue, we propose a graph-based approach that enforces the continuity and connectivity of anatomical topology in medical images. Our method encodes the continuity of shapes as a graph constraint, ensuring that the network’s predictions maintain this continuity. We evaluate our method on three public benchmarks of retinal vessel segmentation and one neuronal structure segmentation benchmark, showing significant improvements in connectivity metrics compared to previous works while getting better or on-par performance on segmentation metrics.

OriginalspracheEnglisch
TitelGraphs in Biomedical Image Analysis, and Overlapped Cell on Tissue Dataset for Histopathology - 5th MICCAI Workshop, GRAIL 2023 and 1st MICCAI Challenge, OCELOT 2023, Held in Conjunction with MICCAI 2023, Proceedings
Redakteure/-innenSeyed-Ahmad Ahmadi, Sérgio Pereira
Herausgeber (Verlag)Springer Science and Business Media Deutschland GmbH
Seiten3-13
Seitenumfang11
ISBN (Print)9783031550874
DOIs
PublikationsstatusVeröffentlicht - 2024
Veranstaltung5th Workshop on GRaphs in biomedicAl Image anaLysis Satellite event at MICCAI, GRAIL 2023 and 1st Cell Detection from Cell-Tissue Interaction challenge in MICCAI, OCELOT 2023 Held in Conjunction with International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2023 - Vancouver, Kanada
Dauer: 8 Okt. 202312 Okt. 2023

Publikationsreihe

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band14373 LNCS
ISSN (Print)0302-9743
ISSN (elektronisch)1611-3349

Konferenz

Konferenz5th Workshop on GRaphs in biomedicAl Image anaLysis Satellite event at MICCAI, GRAIL 2023 and 1st Cell Detection from Cell-Tissue Interaction challenge in MICCAI, OCELOT 2023 Held in Conjunction with International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2023
Land/GebietKanada
OrtVancouver
Zeitraum8/10/2312/10/23

Fingerprint

Untersuchen Sie die Forschungsthemen von „SCOPE: Structural Continuity Preservation for Retinal Vessel Segmentation“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren