Keywords: Segmentation, Machine Learning/Artificial Intelligence, Brain Metastases
Motivation: Brain metastasis treatment monitoring requires precise tumour segmentation, but manual methods are time-intensive and prone to variability. Current methods overlook metabolic and microstructural changes, focusing only on macrostructural components.
Goal(s): Develop an automatic segmentation model that delineates Gd-enhanced, edema, and Nuclear Overhauser Magnetization Transfer ratio (NOE-MT) attenuated sub-regions, incorporating metabolic and microstructural information for improved boundary definition.
Approach: Data from 101 patients consisting of contrast-enhanced T1w, FLAIR, and MTRNOE images, trained a modified U-Net with spatial attention. Performance was evaluated using Dice and boundary F1-scores.
Results: The model achieved reliable segmentation, demonstrating strong boundary delineation and potential for treatment monitoring.
Impact: Introduces the novel Nuclear Overhauser Magnetization Transfer (NOE-MT) sub-region segmentation, providing metabolic and microstructural tumour insights. An automatic model that accurately delineates brain metastasis sub-regions, enhancing boundary definition and reducing observer variability, with implications for improved treatment planning and monitoring.
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.
Keywords