What question did this study set out to answer?

The aim is to develop a more effective framework for brain tumor segmentation by addressing limitations of existing methods.

Hybrid of Handcrafted Features and Learnable Features Using Multimodal Fusion Network for Brain Tumor Segmentation

View Full Paper

Ask AI

Bookmark

Authors

This research demonstrates enhanced brain tumor segmentation in medical imaging, suggesting improved diagnostic capabilities in clinical settings.

The aim is to develop a more effective framework for brain tumor segmentation by addressing limitations of existing methods.
Proposed a lightweight HHL-MFNet framework for brain tumor segmentation.
Utilized layer-level fusion to integrate information from multiple modalities.
Combined handcrafted Discrete Wavelet Transform (DWT) features with a learnable encoder-decoder architecture.
Incorporated Multimodal Fusion (MMF) and Cross-Modal Integration (CIM) modules.
HHL-MFNet outperforms eight other models on BraTS 2020 and BraTS 2021 datasets.
Achieved superior performance metrics including dice score, HDF95, and sensitivity.
Significantly fewer learnable parameters than a competing model with slightly better performance.

Bhattacharjee et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

This research demonstrates enhanced brain tumor segmentation in medical imaging, suggesting improved diagnostic capabilities in clinical settings.