Meeting Banner
Abstract #1756

Focused MRI Segmentation: Leveraging Diffusion Models for Brain Tumor Segmentation in Low-Resolution Areas

Luis Carlos Rivera Monroy1, Tianqi Wang1,2, Vasileios Belagiannis2, and Andreas Maier1
1Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), Erlangen, Germany, 2Electrical-Electronic-Communication Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), Erlangen, Germany

Synopsis

Keywords: AI Diffusion Models, Segmentation

Motivation: Glioblastoma, the most prevalent and aggressive adult brain tumor, presents a therapeutic and monitoring challenge due to its diverse morphology and composition.

Goal(s): This study explores the efficacy of utilizing denoising diffusion models with the widely adopted U-Net architecture for enhanced segmentation performance.

Approach: The proposed framework notably improves the segmentation of the tumor, especially the core. This enhancement facilitates an advanced understanding of complex cases and potentially impacts specialist interventions.

Results: Our findings present promising results for further research into more intricate glioblastoma cases, thereby aiding in developing sophisticated, targeted treatment strategies for this disease.

Impact: This study advances U-Net architecture by integrating denoising diffusion models and specialized loss functions, elevating the precision of low-resolution brain tumor segmentation, with an emphasis on balancing improved accuracy against uncertain predictions.

How to access this content:

For one year after publication, abstracts and videos are only open to registrants of this annual meeting. Registrants should use their existing login information. Non-registrant access can be purchased via the ISMRM E-Library.

After one year, current ISMRM & ISMRT members get free access to both the abstracts and videos. Non-members and non-registrants must purchase access via the ISMRM E-Library.

After two years, the meeting proceedings (abstracts) are opened to the public and require no login information. Videos remain behind password for access by members, registrants and E-Library customers.

Click here for more information on becoming a member.

Keywords