SAMF-Net: Low-Concentration MPI Reconstruction with Segmentation-Assisted Attention and Multi-Harmonic Fusion

Chang Wang; Wenxuan Zou; Guangxing Zhou; Xun Zhang; Yang Jing; Yu An; Weihua Li; Zhenchao Tang; Jie Tian

doi:10.1109/ISBI61048.2026.11516071

SAMF-Net: Low-Concentration MPI Reconstruction with Segmentation-Assisted Attention and Multi-Harmonic Fusion

Chang Wang
, Wenxuan Zou
, Guangxing Zhou
, Xun Zhang
, Yang Jing
, Yu An
, Weihua Li^*
, Zhenchao Tang^*
, Jie Tian^*

^*Corresponding author for this work

School of Engineering Medicine

Beihang University
Peking University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Magnetic Particle Imaging (MPI) is an emerging non-invasive imaging modality for real-time superparamagnetic iron oxide nanoparticles (SPIONs) visualization, with great oncology potential. However, its inherently ill-posed reconstruction suffers severe SNR degradation, artifacts, resolution loss, and quantification inaccuracy at low concentrations. For the ISBI LCR-MPI Challenge, which requires high-fidelity reconstruction from low-concentration free-field line (FFL) MPI data, we propose a novel framework integrating targeted data augmentation, two-stage attention-based reconstruction, segmentation guidance, and multi-harmonic fusion. Our method achieves 0.684 SSIM and 20.32 dB PSNR on the challenge datasets, ranking 4th, and provides a feasible solution for low-to-high concentration MPI image reconstruction.

Original language	English
Title of host publication	ISBI 2026 - 23rd IEEE International Symposium on Biomedical Imaging
Publisher	IEEE Computer Society
ISBN (Electronic)	9798331577636
DOIs	https://doi.org/10.1109/ISBI61048.2026.11516071
State	Published - 2026
Event	23rd IEEE International Symposium on Biomedical Imaging, ISBI 2026 - London, United Kingdom Duration: 8 Apr 2026 → 11 Apr 2026

Publication series

Name	Proceedings - International Symposium on Biomedical Imaging
Volume	2026-April
ISSN (Print)	1945-7928
ISSN (Electronic)	1945-8452

Conference

Conference	23rd IEEE International Symposium on Biomedical Imaging, ISBI 2026
Country/Territory	United Kingdom
City	London
Period	8/04/26 → 11/04/26

Access to Document

10.1109/ISBI61048.2026.11516071

Cite this

Wang, C., Zou, W., Zhou, G., Zhang, X., Jing, Y., An, Y., Li, W., Tang, Z., & Tian, J. (2026). SAMF-Net: Low-Concentration MPI Reconstruction with Segmentation-Assisted Attention and Multi-Harmonic Fusion. In ISBI 2026 - 23rd IEEE International Symposium on Biomedical Imaging (Proceedings - International Symposium on Biomedical Imaging; Vol. 2026-April). IEEE Computer Society. https://doi.org/10.1109/ISBI61048.2026.11516071

@inproceedings{e9f54cc9df5842ba8b9aee5967edb59a,

title = "SAMF-Net: Low-Concentration MPI Reconstruction with Segmentation-Assisted Attention and Multi-Harmonic Fusion",

abstract = "Magnetic Particle Imaging (MPI) is an emerging non-invasive imaging modality for real-time superparamagnetic iron oxide nanoparticles (SPIONs) visualization, with great oncology potential. However, its inherently ill-posed reconstruction suffers severe SNR degradation, artifacts, resolution loss, and quantification inaccuracy at low concentrations. For the ISBI LCR-MPI Challenge, which requires high-fidelity reconstruction from low-concentration free-field line (FFL) MPI data, we propose a novel framework integrating targeted data augmentation, two-stage attention-based reconstruction, segmentation guidance, and multi-harmonic fusion. Our method achieves 0.684 SSIM and 20.32 dB PSNR on the challenge datasets, ranking 4th, and provides a feasible solution for low-to-high concentration MPI image reconstruction.",

author = "Chang Wang and Wenxuan Zou and Guangxing Zhou and Xun Zhang and Yang Jing and Yu An and Weihua Li and Zhenchao Tang and Jie Tian",

note = "Publisher Copyright: {\textcopyright} 2026 IEEE.; 23rd IEEE International Symposium on Biomedical Imaging, ISBI 2026 ; Conference date: 08-04-2026 Through 11-04-2026",

year = "2026",

doi = "10.1109/ISBI61048.2026.11516071",

language = "英语",

series = "Proceedings - International Symposium on Biomedical Imaging",

publisher = "IEEE Computer Society",

booktitle = "ISBI 2026 - 23rd IEEE International Symposium on Biomedical Imaging",

address = "美国",

}

Wang, C, Zou, W, Zhou, G, Zhang, X, Jing, Y, An, Y, Li, W, Tang, Z & Tian, J 2026, SAMF-Net: Low-Concentration MPI Reconstruction with Segmentation-Assisted Attention and Multi-Harmonic Fusion. in ISBI 2026 - 23rd IEEE International Symposium on Biomedical Imaging. Proceedings - International Symposium on Biomedical Imaging, vol. 2026-April, IEEE Computer Society, 23rd IEEE International Symposium on Biomedical Imaging, ISBI 2026, London, United Kingdom, 8/04/26. https://doi.org/10.1109/ISBI61048.2026.11516071

SAMF-Net: Low-Concentration MPI Reconstruction with Segmentation-Assisted Attention and Multi-Harmonic Fusion. / Wang, Chang; Zou, Wenxuan; Zhou, Guangxing et al.
ISBI 2026 - 23rd IEEE International Symposium on Biomedical Imaging. IEEE Computer Society, 2026. (Proceedings - International Symposium on Biomedical Imaging; Vol. 2026-April).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - SAMF-Net

T2 - 23rd IEEE International Symposium on Biomedical Imaging, ISBI 2026

AU - Wang, Chang

AU - Zou, Wenxuan

AU - Zhou, Guangxing

AU - Zhang, Xun

AU - Jing, Yang

AU - An, Yu

AU - Li, Weihua

AU - Tang, Zhenchao

AU - Tian, Jie

PY - 2026

Y1 - 2026

N2 - Magnetic Particle Imaging (MPI) is an emerging non-invasive imaging modality for real-time superparamagnetic iron oxide nanoparticles (SPIONs) visualization, with great oncology potential. However, its inherently ill-posed reconstruction suffers severe SNR degradation, artifacts, resolution loss, and quantification inaccuracy at low concentrations. For the ISBI LCR-MPI Challenge, which requires high-fidelity reconstruction from low-concentration free-field line (FFL) MPI data, we propose a novel framework integrating targeted data augmentation, two-stage attention-based reconstruction, segmentation guidance, and multi-harmonic fusion. Our method achieves 0.684 SSIM and 20.32 dB PSNR on the challenge datasets, ranking 4th, and provides a feasible solution for low-to-high concentration MPI image reconstruction.

AB - Magnetic Particle Imaging (MPI) is an emerging non-invasive imaging modality for real-time superparamagnetic iron oxide nanoparticles (SPIONs) visualization, with great oncology potential. However, its inherently ill-posed reconstruction suffers severe SNR degradation, artifacts, resolution loss, and quantification inaccuracy at low concentrations. For the ISBI LCR-MPI Challenge, which requires high-fidelity reconstruction from low-concentration free-field line (FFL) MPI data, we propose a novel framework integrating targeted data augmentation, two-stage attention-based reconstruction, segmentation guidance, and multi-harmonic fusion. Our method achieves 0.684 SSIM and 20.32 dB PSNR on the challenge datasets, ranking 4th, and provides a feasible solution for low-to-high concentration MPI image reconstruction.

UR - https://www.scopus.com/pages/publications/105041636656

U2 - 10.1109/ISBI61048.2026.11516071

DO - 10.1109/ISBI61048.2026.11516071

M3 - 会议稿件

AN - SCOPUS:105041636656

T3 - Proceedings - International Symposium on Biomedical Imaging

BT - ISBI 2026 - 23rd IEEE International Symposium on Biomedical Imaging

PB - IEEE Computer Society

Y2 - 8 April 2026 through 11 April 2026

ER -

SAMF-Net: Low-Concentration MPI Reconstruction with Segmentation-Assisted Attention and Multi-Harmonic Fusion

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this