PHPA ‐Net: A Paired Hybrid Fusion and Parallel Attention 3D U‐Net for MRI Brain Tumor Segmentation
International Journal of Imaging Systems and Technology · 2026-11 · 36 (6)
DOI: 10.1002/ima.70471 ↗At a glance
Journal article in International Journal of Imaging Systems and Technology, published online on 2026-11-09. Written by Zhibo Liu and 5 co-authors. 42 references. Funded by: National Natural Science Foundation of China.
Abstract (original language)
Accurate brain‐tumor segmentation requires effective integration of complementary MRI information and spatial feature refinement across heterogeneous tumor regions. We developed PHPA‐Net, a three‐dimensional multimodal network that combines paired cross‐modal interaction during encoding with multiscale and directional feature refinement during decoding. Its paired hybrid fusion module constructs residual interactions between FLAIR–T2 and T1–T1ce, feeds them back into modality‐specific streams, and postpones four‐modality integration until the bottleneck and skip pathways. A parallel attention fusion module then combines multireceptive‐field context with input‐dependent triaxial spatial aggregation. Across custom BraTS 2018, BraTS 2019, and BraTS 2021 evaluations, PHPA‐Net showed competitive overlap performance. On the BraTS 2021 split, it achieved a mean Dice of 89.6% and a mean HD95 of 3.8 mm, the best values among the evaluated methods.
Funding
- National Natural Science Foundation of China
Cite
APA
Liu, Z., Wu, X., Hou, B., Wang, Y., Xu, Z., & Zhang, Y. (2026). PHPA ‐Net: A Paired Hybrid Fusion and Parallel Attention 3D U‐Net for MRI Brain Tumor Segmentation. International Journal of Imaging Systems and Technology, 36(6). https://doi.org/10.1002/ima.70471
Vancouver
Liu Z, Wu X, Hou B, Wang Y, Xu Z, Zhang Y. PHPA ‐Net: A Paired Hybrid Fusion and Parallel Attention 3D U‐Net for MRI Brain Tumor Segmentation. International Journal of Imaging Systems and Technology. 2026;36(6). doi:10.1002/ima.70471
BibTeX
@article{Liu2026,
author = {Liu, Zhibo and Wu, Xiaosheng and Hou, Beibei and Wang, Yuwen and Xu, Zhaozhao and Zhang, Yudong},
title = {PHPA ‐Net: A Paired Hybrid Fusion and Parallel Attention 3D U‐Net for MRI Brain Tumor Segmentation},
journal = {International Journal of Imaging Systems and Technology},
year = {2026},
volume = {36},
number = {6},
doi = {10.1002/ima.70471}
}
EUVIMED