Scattering-Based Self-Supervised Learning for Label-Efficient Cardiac Image Segmentation

Basit Öğe Kaydı
dc.contributor.authorAlasu, Serdar
dc.contributor.authorTalu, Muhammed Fatih
dc.date.accessioned2026-04-04T13:31:08Z
dc.date.available2026-04-04T13:31:08Z
dc.date.issued2026
dc.departmentİnönü Üniversitesi
dc.description.abstractDeep learning models based on supervised learning rely heavily on large annotated datasets and particularly in the context of medical image segmentation, the requirement for pixel-level annotations makes the labeling process labor-intensive, time-consuming and expensive. To overcome these limitations, self-supervised learning (SSL) has emerged as a promising alternative that learns generalizable representations from unlabeled data; however, existing SSL frameworks often employ highly parameterized encoders that are computationally expensive and may lack robustness in label-scarce settings. In this work, we propose a scattering-based SSL framework that integrates Wavelet Scattering Networks (WSNs) and Parametric Scattering Networks (PSNs) into a Bootstrap Your Own Latent (BYOL) pretraining pipeline. By replacing the initial stages of the BYOL encoder with fixed or learnable scattering-based front-ends, the proposed method reduces the number of learnable parameters while embedding translation-invariant and small deformation-stable representations into the SSL pipeline. The pretrained encoders are transferred to a U-Net and fine-tuned for cardiac image segmentation on two datasets with different imaging modalities, namely, cardiac cine MRI (ACDC) and cardiac CT (CHD), under varying amounts of labeled data. Experimental results show that scattering-based SSL pretraining consistently improves segmentation performance over random initialization and ImageNet pretraining in low-label regimes, with particularly pronounced gains when only a few labeled patients are available. Notably, the PSN variant achieves improvements of 4.66% and 2.11% in average Dice score over standard BYOL with only 5 and 10 labeled patients, respectively, on the ACDC dataset. These results demonstrate that integrating mathematically grounded scattering representations into SSL pipelines provides a robust and data-efficient initialization strategy for cardiac image segmentation, particularly under limited annotation and domain shift.
dc.identifier.doi10.3390/electronics15030506
dc.identifier.issn2079-9292
dc.identifier.issue3
dc.identifier.scopus2-s2.0-105030072834
dc.identifier.scopusqualityN/A
dc.identifier.urihttps://doi.org/10.3390/electronics15030506
dc.identifier.urihttps://hdl.handle.net/11616/108590
dc.identifier.volume15
dc.identifier.wosWOS:001687755100001
dc.identifier.wosqualityQ2
dc.indekslendigikaynakWeb of Science
dc.indekslendigikaynakScopus
dc.language.isoen
dc.publisherMdpi
dc.relation.ispartofElectronics
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/openAccess
dc.snmzKA_WOS_20250329
dc.subjectself-supervised learning
dc.subjectwavelet scattering networks
dc.subjectparametric scattering networks
dc.subjectlabel-efficient learning
dc.subjectmedical image segmentation
dc.subjectcardiac image segmentation
dc.titleScattering-Based Self-Supervised Learning for Label-Efficient Cardiac Image Segmentation
dc.typeArticle

Dosyalar

Koleksiyon

WoS İndeksli Yayınlar Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu