Communication Dans Un Congrès Année : 2025

BSRPCA : A Simplified Blind Super-Resolved RPCA-Based Approach for Enhancing Blood Flow Estimation

Résumé

Ultrasound Localization Microscopy (ULM) offers great advances compared to conventional ultrasound imaging, enabling the reconstruction of microvascular structures with micrometer-scale precision. Most model-based localization approaches for ULMrelyonasequentialfive-stepframework, which complicates parameter tuning and results in intricate tissue filtering. In addition, insufficient consideration of point spread function (PSF) also affects the final results. End-toend data-driven solutions offer promising results yet require training data and ground-truths. In this study, we introduced a novel approach, based on computational super-resolution referred to as BSRPCA- Blind Super-Resolved Robust Principal Component Analysis, to address these challenges. Our method replaces the first three steps of the ULM process with a single super-resolution step. Experimental results on both in silico and in vivo datasets have proven the effectiveness of our method compared to the other benchmarks.

Fichier principal
Vignette du fichier
ICIP_BSRPCA_2025_final_version.pdf (858 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05008877 , version 1 (22-05-2025)

Licence

Identifiants

  • HAL Id : hal-05008877 , version 1

Citer

Xuan-Hieu Le, Jean-Pierre Remeniéras, Denis Kouamé, Duong-Hung Pham. BSRPCA : A Simplified Blind Super-Resolved RPCA-Based Approach for Enhancing Blood Flow Estimation. 2025 IEEE International Conference on Image Processing (ICIP 2025), Sep 2025, Anchorage, AL, United States. ⟨hal-05008877⟩
1181 Consultations
481 Téléchargements

Partager

  • More