Continuous implicit neural representation for arbitrary super-resolution of system matrix in magnetic particle imaging DOI
Zhongrong Miao, Liwen Zhang, Jie Tian

et al.

Physics in Medicine and Biology, Journal Year: 2024, Volume and Issue: 70(4), P. 045012 - 045012

Published: Dec. 30, 2024

Abstract Objective . Magnetic particle imaging (MPI) is a novel technique that uses magnetic fields to detect tracer materials consisting of nanoparticles. System matrix (SM) based image reconstruction essential for achieving high quality in MPI. However, the time-consuming SM calibrations need be repeated whenever field’s or nanoparticle’s characteristics change. Accelerating this calibration process therefore crucial. The most common acceleration approach involves undersampling during procedure, followed by super-resolution methods recover high-resolution SM. these typically require separate training multiple models different ratios, leading increased storage and time costs. Approach We propose an arbitrary-scale method on continuous implicit neural representation (INR). Using INR, modeled as function space, enabling sampling at densities. A cross-frequency encoder implemented share frequency information analyze contextual relationships, resulting more intelligent efficient strategy. Convolutional networks (CNNs) are utilized learn optimize grid leveraging advantage CNNs learning local feature associations considering surrounding comprehensively. Main results Experimental OpenMPI demonstrate our outperforms existing enables any scale with single model. proposed achieves accuracy efficiency recovery, even rates. Significance significantly reduces costs associated calibration, making it practical real-world applications. By model, enhances flexibility MPI systems, paving way widespread adoption technology.

Language: Английский

Autonomous robotic ultrasound scanning system: a key to enhancing image analysis reproducibility and observer consistency in ultrasound imaging DOI Creative Commons
Xin-Xin Lin,

Ming‐De Li,

Si‐Min Ruan

et al.

Frontiers in Robotics and AI, Journal Year: 2025, Volume and Issue: 12

Published: Feb. 5, 2025

Purpose This study aims to develop an autonomous robotic ultrasound scanning system (auto-RUSS) pipeline, comparing its reproducibility and observer consistency in image analysis with physicians of varying levels expertise. Design/methodology/approach An auto-RUSS was engineered using a 7-degree-of-freedom arm, real-time regulation based on force control visual servoing. Two phantoms were employed for the human-machine comparative experiment, involving three groups: auto-RUSS, non-expert (4 junior physicians), expert senior physicians). setup enabled comprehensive assessment contact force, acquisition, measurement AI-assisted classification. Radiological feature variability measured coefficient variation (COV), while performance assessments utilized mean standard deviation (SD). Findings The had potential reduce operator-dependent examinations, offering enhanced repeatability across multiple dimensions including probe images measurement, diagnostic model performance. Originality/value In this paper, pipeline proposed. Through comparison experiments, shown effectively improve minimize human-induced variability.

Language: Английский

Citations

0
A novel computer-aided energy decision-making system improves patient treatment by microwave ablation of thyroid nodule DOI Creative Commons
Rui Du, Rui Wang, Hu Xu

et al.

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 188, P. 109823 - 109823

Published: Feb. 21, 2025

Language: Английский

Citations

0
Integrating permutation feature importance with conformal prediction for robust Explainable Artificial Intelligence in predictive process monitoring DOI Creative Commons
Nijat Mehdiyev,

Maxim Majlatow,

Peter Fettke

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 149, P. 110363 - 110363

Published: March 15, 2025

Language: Английский

Citations

0
Shape-intensity-guided U-net for medical image segmentation DOI
Wenhui Dong, Bo Du, Yongchao Xu

et al.

Neurocomputing, Journal Year: 2024, Volume and Issue: 610, P. 128534 - 128534

Published: Sept. 3, 2024

Language: Английский

Citations

3
Bilateral-Aware and Multi-Scale Region Guided U-Net for precise breast lesion segmentation in ultrasound images DOI
Yangyang Li,

Xintong Hou,

Xuanting Hao

et al.

Neurocomputing, Journal Year: 2025, Volume and Issue: unknown, P. 129775 - 129775

Published: Feb. 1, 2025

Language: Английский

Citations

0
Lesion attention guided neural network for contrast-enhanced mammography-based biomarker status prediction in breast cancer DOI
N. Qian, Wei Jiang, Xiaoqian Wu

et al.

Computer Methods and Programs in Biomedicine, Journal Year: 2024, Volume and Issue: 250, P. 108194 - 108194

Published: April 22, 2024

Language: Английский

Citations

3
Segmentation-assisted hierarchical constrained state space approach for robust carotid artery wall motion measurement DOI

Jinhui Wu,

Heye Zhang, Xiujian Liu

et al.

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 254, P. 124377 - 124377

Published: May 31, 2024

Language: Английский

Citations

1
GLCV-NET: An automatic diagnosis system for advanced liver fibrosis using global-local cross view in B-mode ultrasound images DOI

Bianzhe Wu,

Ze-Rong Huang,

Jinglin Liang

et al.

Computer Methods and Programs in Biomedicine, Journal Year: 2024, Volume and Issue: 257, P. 108440 - 108440

Published: Sept. 26, 2024

Language: Английский

Citations

1
TGNet: tensor-based graph convolutional networks for multimodal brain network analysis DOI Creative Commons
Zhaoming Kong, Rong Zhou,

Xinwei Luo

et al.

BioData Mining, Journal Year: 2024, Volume and Issue: 17(1)

Published: Dec. 6, 2024

Multimodal brain network analysis enables a comprehensive understanding of neurological disorders by integrating information from multiple neuroimaging modalities. However, existing methods often struggle to effectively model the complex structures multimodal networks. In this paper, we propose novel tensor-based graph convolutional (TGNet) framework that combines tensor decomposition with multi-layer GCNs capture both homogeneity and intricate We evaluate TGNet on four datasets—HIV, Bipolar Disorder (BP), Parkinson's Disease (PPMI), Alzheimer's (ADNI)—demonstrating it significantly outperforms for disease classification tasks, particularly in scenarios limited sample sizes. The robustness effectiveness highlight its potential advancing analysis. code is available at https://github.com/rongzhou7/TGNet .

Language: Английский

Citations

1
Multiscale Feature Fusion Method for Liver Cirrhosis Classification DOI
Shanshan Wang, Ling Jian, Kaiyan Li

et al.

International Journal of Imaging Systems and Technology, Journal Year: 2024, Volume and Issue: 34(4)

Published: July 1, 2024

ABSTRACT Liver cirrhosis is one of the most common liver diseases in world, posing a threat to people's daily lives. In advanced stages, can lead severe symptoms and complications, making early detection treatment crucial. This study aims address this critical healthcare challenge by improving accuracy classification using ultrasound imaging, thereby assisting medical professionals diagnosis intervention. article proposes new multiscale feature fusion network model (MSFNet), which uses extraction module capture features from images. approach enables neural utilize richer information accurately classify stage cirrhosis. addition, loss function proposed solve class imbalance problem datasets, makes pay more attention samples that are difficult improves performance model. The effectiveness MSFNet was evaluated images 61 subjects. Experimental results demonstrate our method achieves high accuracy, with 98.08% on convex array datasets 97.60% linear datasets. Our early, middle, late very accurately. It provides valuable insights for clinical may be helpful rehabilitation patients.

Language: Английский

Citations

1