Skip to content

Machine Learning for Biomedical Imaging Benchmarking Self-Supervised Learning Methods for Accelerated MRI Reconstruction

· 0 citations · 125 references

TL;DR

SSIBench is presented, a modular and flexible comparison framework to unify and thoroughly benchmark S elf-S upervised I maging methods (SSI) without GT, and is provided the extensible benchmark and open-source reimplementations of all methods at https://github.com/Andrewwango/ssibench.

View source

Similar papers

Preprint Oct 2026

Beyond Training from Scratch: Foundation Models for Data-Efficient and Generalizable Cardiac MRI Reconstruction

This work proposes a reconstruction framework that integrates frozen and parameter-efficiently adapted visual encoders, including CLIP, BiomedCLIP, and DINOv2, within a transformer-based reconstruction architecture and evaluates performance under limited supervision and cross-dataset transfer, showing that foundation m...

Anam Hashmi, Mayug Maniparambil, Julia Dietlmeier et al. · 0 citations
Open access Sep 2026

Uncertainty-weighted dual-domain consistency for self-supervised accelerated MRI reconstruction

A novel dual-domain self-supervised learning (Dual-SSL) framework that rigorously enforces physical consistency in both image and k-space domains is proposed, effectively bridging the gap between self-supervised and supervised MRI reconstruction.

Xin-Zhen Li, Jin-Hong Huang, Gen-Jiao Zhou et al. · 0 citations
Preprint Sep 2026

Physics-informed denoising method for image reconstruction in quantitative low-field MRI

The proposed modular unrolled end-to-end deep learning method and its underlying method offer an efficient and flexible solution to denoise low-field MR data and make quantitative low-field MRI a feasible diagnostic tool for clinical applications.

Catarina Redshaw Kranich, C. Prieto, Christoph Kolbitsch et al. · 0 citations
Sep 2026

A Reliability-Focused Deep Learning Framework for Medical Image Analysis

Modern Medical Imaging methods like X-ray, Ultrasound, CT scan, and MRI are important technologies used to diagnose various diseases since they help healthcare professionals identify the presence of specific structures inside patient's bodies. However, these modern techniques may produce unreliable results that lead to...

Atik Naikwade, Sarvesh Chowdary · 0 citations
Open access Jan 2026

Benchmarking deep learning-based reconstruction methods for photoacoustic computed tomography with clinically relevant synthetic datasets

A benchmarking framework is proposed that provides open-source, anatomically plausible synthetic datasets and evaluation strategies for DL-based acoustic inversion methods in PACT and enables systematic comparisons of DL-based acoustic inversion methods for 2D PACT.

Pan-Pan Chen, Seonyeong Park, Gangwon Jeong et al. · 2 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.