Sensitivity of Quantitative Susceptibility Mapping for Clinical Research in Deep Gray Matter
Year: 2025
Type: article
Abstract: Quantitative susceptibility mapping (QSM) is an advanced MRI technique for assessing iron, calcium, and myelin tissue levels based on magnetic susceptibility. QSM consists of multiple processing steps, with various choices for each step. While QSM is increasingly applied in neurodegenerative disease research, its reproducibility and sensitivity in detecting susceptibility changes across groups or over time, which underpin the interpretation of clinical outcomes, have not been thoroughly quantified. This study aimed to evaluate how choices in background field removal (BFR), dipole inversion algorithms, and anatomical referencing impact the detection of changes in deep gray matter susceptibility. We used aging-related changes in brain iron, established in earlier foundational studies, as a surrogate model to test the sensitivity and reproducibility of 378 different QSM pipelines toward the detection of longitudinal susceptibility changes in a clinical setting. We used 10-year follow-up data and scan-rescan data of healthy adults scanned at 3T. Our results demonstrated high variability in the sensitivity of QSM pipelines toward detecting susceptibility changes. While most pipelines detected the same over-time changes, the choice of the BFR algorithm and the referencing strategy influenced reproducibility error and sensitivity substantially. Notably, pipelines using RESHARP with AMP-PE, HEIDI, or LSQR inversion showed the highest overall sensitivity. The findings suggest a strong impact of algorithmic choices in QSM processing on the ability to detect physiological changes in the brain. Careful consideration should be given to the pipeline configuration for reliable clinical outcomes.
Source: Human Brain Mapping
DOI: 10.1002/hbm.70187
Authors: Fahad Salman(Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Visualization, Writing – original draft, Writing – review & editing), Abhisri Ramesh(Methodology, Software), Thomas Jochmann(Methodology, Software, Writing – review & editing), Mirjam Prayer(Methodology), Ademola Adegbemigun(Methodology), Jack A. Reeves(Formal analysis, Investigation, Writing – review & editing), Gregory E. Wilding(Methodology, Supervision), Junghun Cho(Formal analysis, Investigation, Writing – review & editing), Dejan Jakimovski(Data curation), Niels Bergsland(Conceptualization, Data curation, Resources, Software, Writing – review & editing), Michael G. Dwyer(Conceptualization, Data curation, Resources, Software), Robert Zivadinov(Conceptualization, Data curation, Resources, Writing – review & editing), Ferdinand Schweser(Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Writing – original draft, Writing – review & editing)
Language: English
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,520
Funders: National Science Foundation, Emory University, European Commission, Deutscher Akademischer Austauschdienst, Bundesministerium für Bildung und Forschung, Seoul National University, Pontificia Universidad Católica de Valparaíso, National Institutes of Health, University at Buffalo, National Institute of Neurological Disorders and Stroke, National Center for Advancing Translational Sciences, NextGenerationEU, German Academic Exchange Service
Awards: Acquisition of High Performance Data and Computing Infrastructure to Advance Biomedical Research, Verbundprojekt: Anonymisierung persönlicher Gesundheitsdaten durch Erzeugung digitaler Avatare in Medizin und Pflege - AVATAR -, G2327092927, G3495744386, Deep Gray Matter Iron and Disease Progression in Multiple Sclerosis, University of Buffalo Clinical and Translational Science Institute, MRI: Acquisition of High Performance Computing Infrastructure to Support Computational and Data-Enabled Science and Engineering
OpenAlex ID: W4409734697
