Abstract #5254

Advancing Vascular Segmentation in Ferumoxytol-enhanced MRA: A Comparative Study of nnUNet and SegMamba

Siyue Li¹, Takegawa Yoshida¹, Kim-Lien Nguyen^1,2,3,4, J. Paul Finn^1,3, and Xiaodong Zhong^1,3

¹Department of Radiological Sciences, University of California Los Angeles, Los Angeles, CA, United States, ²Division of Cardiology, VA Greater Los Angeles Healthcare System, Los Angeles, CA, United States, ³Physics and Biology in Medicine Graduate Program, University of California, Los Angeles, CA, United States, ⁴Division of Cardiology, University of California Los Angeles, Los Angeles, CA, United States

Synopsis

Keywords: Vascular/Vessel Wall, Segmentation

Motivation: Vascular segmentation in ferumoxytol-enhanced MR angiography (FE-MRA) is essential for diagnosing vascular diseases, but traditional methods often require time-consuming manual intervention. Deep learning models like nnUNet and SegMamba are promising to improve automation and diagnostic accuracy.

Goal(s): This study compares the performance of nnUNet and SegMamba for automated vascular segmentation in FE-MRA.

Approach: We evaluated both models with an FE-MRA dataset of 12 patients, using the Dice Similarity Coefficient to assess the segmentation accuracy.

Results: nnUNet outperformed SegMamba, demonstrating a DSC of 0.853 and 0.828 when trained with data from 8 and 4 patients respectively.

Impact: This study explores automated vascular segmentation in FE-MRA and demonstrates that nnUNet outperforms SegMamba, offering a more reliable and automated approach.

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.