This is a page that curates AI-related papers published worldwide. All content here is summarized using Google Gemini and operated on a non-profit basis. Copyright for each paper belongs to the authors and their institutions; please make sure to credit the source when sharing.
This paper presents SRMA-Mamba, a novel medical image analysis model for accurate diagnosis and treatment of liver cirrhosis. To overcome the limitations of existing methods, which fail to fully utilize the spatial anatomical details of MRI data, SRMA-Mamba integrates a Mamba-based network with a spatial anatomy-based Mamba module (SABMamba) and a spatial backward attention module (SRMA). SABMamba performs selective Mamba scans within cirrhotic tissue and combines sagittal, coronal, and axial anatomical information to construct a comprehensive spatial contextual representation. SRMA utilizes a coarse segmentation map and hierarchical encoding features to progressively enhance cirrhotic details. Experimental results demonstrate that SRMA-Mamba outperforms existing state-of-the-art methods in 3D pathological liver segmentation.
Takeaways, Limitations
•
Takeaways :
◦
We present a novel model, SRMA-Mamba, which outperforms existing methods in segmenting 3D pathological liver structures in cirrhosis.
◦
Effectively utilizing spatial anatomical information of MRI data to improve accuracy.
◦
Ensure reproducibility and extensibility through open code.
•
Limitations :
◦
Further validation of the generalization performance of the proposed model is needed.
◦
Applicability studies for various types of liver diseases are needed.
◦
Additional validation and clinical data are needed for clinical application.
