Shape-Aware Cell Segmentation with Global Cues

J 2026

Shape-Aware Cell Segmentation with Global Cues

HRADECKÁ, Lucia a Martin MAŠKA

Základní údaje

Originální název

Shape-Aware Cell Segmentation with Global Cues

Autoři

HRADECKÁ, Lucia a Martin MAŠKA

Vydání

IEEE SIGNAL PROCESSING LETTERS, UNITED STATES, IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2026, 1070-9908

Další údaje

Jazyk

angličtina

Typ výsledku

Článek v odborném periodiku

Obor

10201 Computer sciences, information science, bioinformatics

Stát vydavatele

Spojené státy

Utajení

není předmětem státního či obchodního tajemství

Odkazy

URL

Impakt faktor

Impact factor: 3.900 v roce 2024

Označené pro přenos do RIV

Ano

Organizační jednotka

Fakulta informatiky

Klíčová slova anglicky

Cell segmentation; deep learning; convolutional neural network; implicit active contours; shape-related cues

Štítky

cbia-web

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 7. 9. 2026 12:22, Mgr. Lucia Hradecká

Anotace

V originále

While convolutional neural networks are still widely used in bioimage segmentation tasks, they suffer from an inherent issue: lack of shape awareness and long-range context due to processing image data exclusively using local operations. In response, we present an architecture that joins CNNs with implicit active contours, deliberately designed to perform segmentation with global and shape-related cues at sub-pixel accuracy. Quantitative validation using three cell imaging modalities and three complementary measures shows that our method is superior to its CNN- and transformer-based competitors as well as a pretrained generalist Cellpose-SAM model: our method decreases boundary localization errors by as much as 82% while achieving reliable overall performance.

Návaznosti

CZ.02.01.01/00/23_015/0008205, interní kód MU

Název: Modernizace VVI Czech-BioImaging

Investor: Ministerstvo školství, mládeže a tělovýchovy ČR, Modernizace VVI Czech-BioImaging, Priorita 1 - Výzkum a vývoj

LM2023050, projekt VaV

Název: Národní infrastruktura pro biologické a medicínské zobrazování

Investor: Ministerstvo školství, mládeže a tělovýchovy ČR, Czech BioImaging: Národní výzkumná infrastruktura pro biologické a medicínské zobrazování

MUNI/A/1784/2025, interní kód MU

Název: Modelování, analýza a verifikace (2026)

Investor: Masarykova univerzita, Modelování, analýza a verifikace (2026)

Citovat

HRADECKÁ, Lucia a Martin MAŠKA. Shape-Aware Cell Segmentation with Global Cues. IEEE SIGNAL PROCESSING LETTERS. UNITED STATES: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2026, roč. 33, č. 2026, s. 3277-3281. ISSN 1070-9908. Dostupné z: https://doi.org/10.1109/LSP.2026.3721049.

@article{2584461,
   author = {Hradecká, Lucia and Maška, Martin},
   article_location = {UNITED STATES},
   article_number = {2026},
   doi = {https://doi.org/10.1109/LSP.2026.3721049},
   keywords = {Cell segmentation; deep learning; convolutional neural network; implicit active contours; shape-related cues},
   language = {eng},
   issn = {1070-9908},
   journal = {IEEE SIGNAL PROCESSING LETTERS},
   title = {Shape-Aware Cell Segmentation with Global Cues},
   url = {https://doi.org/10.1109/LSP.2026.3721049},
   volume = {33},
   year = {2026}
}

TY  - JOUR
ID  - 2584461
AU  - Hradecká, Lucia - Maška, Martin
PY  - 2026
TI  - Shape-Aware Cell Segmentation with Global Cues
JF  - IEEE SIGNAL PROCESSING LETTERS
VL  - 33
IS  - 2026
SP  - 3277-3281
EP  - 3277-3281
PB  - IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
SN  - 10709908
KW  - Cell segmentation
KW  - deep learning
KW  - convolutional neural network
KW  - implicit active contours
KW  - shape-related cues
UR  - https://doi.org/10.1109/LSP.2026.3721049
N2  - While convolutional neural networks are still widely used in bioimage segmentation tasks, they suffer from an inherent issue: lack of shape awareness and long-range context due to processing image data exclusively using local operations. In response, we present an architecture that joins CNNs with implicit active contours, deliberately designed to perform segmentation with global and shape-related cues at sub-pixel accuracy. Quantitative validation using three cell imaging modalities and three complementary measures shows that our method is superior to its CNN- and transformer-based competitors as well as a pretrained generalist Cellpose-SAM model: our method decreases boundary localization errors by as much as 82% while achieving reliable overall performance.
ER  -

HRADECKÁ, Lucia a Martin MAŠKA. Shape-Aware Cell Segmentation with Global Cues. \textit{IEEE SIGNAL PROCESSING LETTERS}. UNITED STATES: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2026, roč.~33, č.~2026, s.~3277-3281. ISSN~1070-9908. Dostupné z: https://doi.org/10.1109/LSP.2026.3721049.

Přehled o publikaci