Synthesis of Double Inversion Recovery-like Images for Superior Contrast and Multiple Sclerosis Lesion Visibility
Motivation: Double Inversion Recovery (DIR) can enhance lesion detection, particularly in multiple sclerosis (MS) where juxtacortical and infratentorialplaques are challenging to detect with conventional techniques. Goal(s): This study evaluates DIR-like images(Syn-DIR) generated from T1 and T2 FLAIR for brain structure and lesion visibility Approach: The study evaluates 30 subjects to assess quantitative segmentation, volumetric consistency, and lesion-to-white-matter contrast. A second longitudinal study cohort assessed 20 additional subjects scanned at two time points to evaluate performance consistency over time. Results: Syn-DIR demonstrated high structural agreement to the reference set and showed significantly higher lesion-to-white matter contrast, with stable gray-to-white-matter contrast across time points. Impact: Syn-DIR provides superior lesion visibility and consistent brain structure representation. The strong agreement in quantitative lesion and regional brain volumes demonstrates its robustness for clinical and research applications.
Authors
- Xiang Lei (ORCID: https://orcid.org/0000-0002-1143-0968)
- Lawrence Tanenbaum (ORCID: https://orcid.org/0000-0003-3326-3192)
- Ajit Shankaranarayanan (ORCID: https://orcid.org/0000-0001-9395-8768)
- Craig B. Arnold (ORCID: https://orcid.org/0000-0002-0293-5411)
- Suzie Bash (ORCID: https://orcid.org/0000-0002-2572-1010)
- Long Wang (ORCID: https://orcid.org/0000-0001-5600-8157)
- Zechen Zhou (ORCID: https://orcid.org/0000-0001-8149-7546)
Publication Details
- Journal
- Proceedings on CD-ROM - International Society for Magnetic Resonance in Medicine. Scientific Meeting and Exhibition/Proceedings of the International Society for Magnetic Resonance in Medicine, Scientific Meeting and Exhibition
- Published
- 2025-09-16
- DOI
- https://doi.org/10.58530/2025/3587
- Primary Topic
- Brain Tumor Detection and Classification
- Type
- article
- Field-Weighted Citation Impact
- 0.00