arXiv subjects

Zixin Hu

Publications and source records attributed to Zixin Hu.

15 recordsLinked to original sources

arXiv2026

MetaStructAtlas: A Grounded 3D Vision-Language Dataset and Benchmark for Functional and Structural Reasoning in Whole-Body PET/CT

The joint interpretation of metabolic function and anatomical structure is essential for clinical diagnosis in whole-body PET/CT. Although recent advances in 3D medical vision-language models have demonstrated remarkable progress, current efforts are limited to regional CT imaging, leaving a critical void in comprehensive whole-body PET/CT analysis. In this work, we introduce MetaStructAtlas, a large-scale dataset for grounded whole-body PET/CT interpretation that synthesizes multimodal imaging with integrated anatomical, metabolic, and semantic annotations. MetaStructAtlas provides 490 co-registered 3D PET and CT volumes with 50,470 organ-level segmentation masks and grounded radiology reports. To facilitate interactive reasoning, we further developed MetaStructVQA, a standardized 3D grounded visual question-answering benchmark containing 100,565 QA pairs. This framework explicitly links diagnostic queries to visual evidence across modalities, encompassing anatomical, morphological, and metabolic characteristics. Finally, we evaluate state-of-the-art 3D medical VLMs on MetaStructVQA, establishing a robust foundation for multimodal representation learning and integrated whole-body reasoning in nuclear medicine.

cs.CV↗

arXiv2026

SemiSAM-O1: Pushing the Boundary of Annotation-Efficient Medical Image Segmentation with Generalist Knowledge Fusion

Semi-supervised learning (SSL) has become a promising solution to alleviate the annotation burden of deep learning-based medical image segmentation models. While recent advances in foundation model-driven SSL have pushed the boundary to extremely limited annotation scenarios, they fail to maintain robust competitive performance in complex imaging modalities. In this paper, we propose SemiSAM-O1, an annotation-efficient framework using only one annotated template case for segmentation. SemiSAM-O1 extends the specialist-generalist collaborative learning framework to the extreme one-label setting by fully exploiting the foundation model's feature representation capability beyond its prompting interface. SemiSAM-O1 operates in two stages. In the first stage, the foundation model's encoder extracts dense features from all volumes, and class prototypes derived from the single annotated template are propagated to the unlabeled pool via feature similarity to produce coarse initial pseudo-labels. In the second stage, an iterative training-and-refinement loop progressively improves both the segmentation model and the pseudo-labels over multiple rounds, where each round trains the model from scratch on current pseudo-labels and generates updated predictions with voxel-wise uncertainty estimates. An uncertainty-guided refinement step further leverages the foundation model's global feature space to correct high-uncertainty regions by aggregating labels from their most similar confident neighbors, establishing a virtuous cycle of mutual improvement. Extensive experiments on a wide range of segmentation tasks across different modalities and anatomical targets demonstrate that SemiSAM-O1 significantly narrows the performance gap between one-label semi-supervised learning and full supervision, while significantly reducing the computational overhead of online foundation model inference.

cs.CV↗

arXiv2026

A Polynomial Architecture-Attribution Co-Design Framework for Exact Aumann-Shapley Attribution in GNNs

We study feature-level and node-level explanations for graph neural networks (GNNs) through the lens of Aumann-Shapley attribution. Path-integral methods such as Integrated Gradients provide an axiomatic formulation of attribution, but their practical use in deep GNNs typically relies on finite-sample numerical approximations to the path integral, requiring a trade-off between quadrature error and computational cost. This paper proposes APEX, a model-attribution co-design framework that makes the attribution integral exactly computable under a polynomial GNN architecture. The key component is PolyGIN, a GIN-style graph network whose message-passing, normalization, and transformation operations preserve a bounded multivariate polynomial form for scalar model scores, such as pre-softmax logits. We show that, for a PolyGIN with $L$ polynomial transformation blocks, the derivative along the attribution path has degree at most $2^L-1$. Therefore, Gauss--Legendre quadrature can evaluate the Aumann--Shapley path integral exactly, up to floating-point precision, with $2^{L-1}$ deterministic evaluation points. The resulting attributions can be computed at the feature level and then aggregated into node-level scores while preserving completeness. Experiments on synthetic and real-world graph benchmarks show that PolyGIN maintains competitive predictive performance, while the complete APEX framework achieves higher attribution fidelity than the compared baselines and substantially reduces the number of evaluations required for path integration.

cs.LG↗

arXiv2026

Universality Reconsidered: Rethinking the Validation of Foundation Models for General-Purpose 3D Medical Segmentation

Foundation models have emerged as a transformative paradigm in 3D medical imaging, with the promise of unified quantitative analysis across diverse targets and imaging modalities. Yet the prevailing conception of universality remains incomplete. Current models are predominantly developed and evaluated on datasets largely concentrated around a limited set of imaging modalities and anatomical regions. In this Perspective, we evaluate representative 3D segmentation foundation models using paired whole-body structural and functional imaging data. Our analysis reveals a substantial gap between benchmark-reported performance and real-world generalization, with marked degradation on previously unseen data and particularly severe failures on functional imaging modalities. These findings suggest that current foundation models remain far from achieving true universality. We argue that progress requires not only scaling models and datasets, but also a reconsideration of how universality is defined and validated, extending evaluation beyond regional structural benchmarks toward whole-body structural and functional imaging. Our observations highlight the need to distinguish benchmark success from genuine clinical generalization. Bridging this gap will be essential for translating foundation models from controlled evaluation settings to real-world medical practice.

cs.CV↗

arXiv2026

Developing Foundation Models for Universal Segmentation from 3D Whole-Body Positron Emission Tomography

Positron emission tomography (PET) is a key nuclear medicine imaging modality that visualizes radiotracer distributions to quantify in vivo physiological and metabolic processes, playing an irreplaceable role in disease management. Despite its clinical importance, the development of deep learning models for quantitative PET image analysis remains severely limited, driven by both the inherent segmentation challenge from PET's paucity of anatomical contrast and the high costs of data acquisition and annotation. To bridge this gap, we develop generalist foundational models for universal segmentation from 3D whole-body PET imaging. We first build the largest and most comprehensive PET dataset to date, comprising 11041 3D whole-body PET scans with 59831 segmentation masks for model development. Based on this dataset, we present SegAnyPET, an innovative foundational model with general-purpose applicability to diverse segmentation tasks. Built on a 3D architecture with a prompt engineering strategy for mask generation, SegAnyPET enables universal and scalable organ and lesion segmentation, supports efficient human correction with minimal effort, and enables a clinical human-in-the-loop workflow. Extensive evaluations on multi-center, multi-tracer, multi-disease datasets demonstrate that SegAnyPET achieves strong zero-shot performance across a wide range of segmentation tasks, highlighting its potential to advance the clinical applications of molecular imaging.

cs.CV↗

arXiv2026

FSX: Message Flow Sensitivity Enhanced Structural Explainer for Graph Neural Networks

Despite the widespread success of Graph Neural Networks (GNNs), understanding the reasons behind their specific predictions remains challenging. Existing explainability methods face a trade-off that gradient-based approaches are computationally efficient but often ignore structural interactions, while game-theoretic techniques capture interactions at the cost of high computational overhead and potential deviation from the model's true reasoning path. To address this gap, we propose FSX (Message Flow Sensitivity Enhanced Structural Explainer), a novel hybrid framework that synergistically combines the internal message flows of the model with a cooperative game approach applied to the external graph data. FSX first identifies critical message flows via a novel flow-sensitivity analysis: during a single forward pass, it simulates localized node perturbations and measures the resulting changes in message flow intensities. These sensitivity-ranked flows are then projected onto the input graph to define compact, semantically meaningful subgraphs. Within each subgraph, a flow-aware cooperative game is conducted, where node contributions are evaluated fairly through a Shapley-like value that incorporates both node-feature importance and their roles in sustaining or destabilizing the identified critical flows. Extensive evaluation across multiple datasets and GNN architectures demonstrates that FSX achieves superior explanation fidelity with significantly reduced runtime, while providing unprecedented insights into the structural logic underlying model predictions--specifically, how important sub-structures exert influence by governing the stability of key internal computational pathways.

cs.LG↗

arXiv2025

PET2Rep: Towards Vision-Language Model-Drived Automated Radiology Report Generation for Positron Emission Tomography

Positron emission tomography (PET) is a cornerstone of modern oncologic and neurologic imaging, distinguished by its unique ability to illuminate dynamic metabolic processes that transcend the anatomical focus of traditional imaging technologies. Radiology reports are essential for clinical decision making, yet their manual creation is labor-intensive and time-consuming. Recent advancements of vision-language models (VLMs) have shown strong potential in medical applications, presenting a promising avenue for automating report generation. However, existing applications of VLMs in the medical domain have predominantly focused on structural imaging modalities, while the unique characteristics of molecular PET imaging have largely been overlooked. To bridge the gap, we introduce PET2Rep, a large-scale comprehensive benchmark for evaluation of general and medical VLMs for radiology report generation for PET images. PET2Rep stands out as the first dedicated dataset for PET report generation with metabolic information, uniquely capturing whole-body image-report pairs that cover dozens of organs to fill the critical gap in existing benchmarks and mirror real-world clinical comprehensiveness. In addition to widely recognized natural language generation metrics, we introduce a series of clinical efficacy metrics to evaluate the quality of radiotracer uptake pattern description in key organs in generated reports. We conduct a head-to-head comparison of 30 cutting-edge general-purpose and medical-specialized VLMs. The results show that the current state-of-the-art VLMs perform poorly on PET report generation task, falling considerably short of fulfilling practical needs. Moreover, we identify several key insufficiency that need to be addressed to advance the development in medical applications.

eess.IV↗

arXiv2025

Efficient Network Automatic Relevance Determination

We propose Network Automatic Relevance Determination (NARD), an extension of ARD for linearly probabilistic models, to simultaneously model sparse relationships between inputs $X \in \mathbb R^{d \times N}$ and outputs $Y \in \mathbb R^{m \times N}$, while capturing the correlation structure among the $Y$. NARD employs a matrix normal prior which contains a sparsity-inducing parameter to identify and discard irrelevant features, thereby promoting sparsity in the model. Algorithmically, it iteratively updates both the precision matrix and the relationship between $Y$ and the refined inputs. To mitigate the computational inefficiencies of the $\mathcal O(m^3 + d^3)$ cost per iteration, we introduce Sequential NARD, which evaluates features sequentially, and a Surrogate Function Method, leveraging an efficient approximation of the marginal likelihood and simplifying the calculation of determinant and inverse of an intermediate matrix. Combining the Sequential update with the Surrogate Function method further reduces computational costs. The computational complexity per iteration for these three methods is reduced to $\mathcal O(m^3+p^3)$, $\mathcal O(m^3 + d^2)$, $\mathcal O(m^3+p^2)$, respectively, where $p \ll d$ is the final number of features in the model. Our methods demonstrate significant improvements in computational efficiency with comparable performance on both synthetic and real-world datasets.

cs.AI↗

arXiv2024

FuXi-S2S: A machine learning model that outperforms conventional global subseasonal forecast models

Skillful subseasonal forecasts are crucial for various sectors of society but pose a grand scientific challenge. Recently, machine learning based weather forecasting models outperform the most successful numerical weather predictions generated by the European Centre for Medium-Range Weather Forecasts (ECMWF), but have not yet surpassed conventional models at subseasonal timescales. This paper introduces FuXi Subseasonal-to-Seasonal (FuXi-S2S), a machine learning model that provides global daily mean forecasts up to 42 days, encompassing five upper-air atmospheric variables at 13 pressure levels and 11 surface variables. FuXi-S2S, trained on 72 years of daily statistics from ECMWF ERA5 reanalysis data, outperforms the ECMWF's state-of-the-art Subseasonal-to-Seasonal model in ensemble mean and ensemble forecasts for total precipitation and outgoing longwave radiation, notably enhancing global precipitation forecast. The improved performance of FuXi-S2S can be primarily attributed to its superior capability to capture forecast uncertainty and accurately predict the Madden-Julian Oscillation (MJO), extending the skillful MJO prediction from 30 days to 36 days. Moreover, FuXi-S2S not only captures realistic teleconnections associated with the MJO, but also emerges as a valuable tool for discovering precursor signals, offering researchers insights and potentially establishing a new paradigm in Earth system science research.

physics.ao-ph↗

arXiv2024

The association of domain-specific physical activity and sedentary activity with stroke: A prospective cohort study

Background The incidence of stroke places a heavy burden on both society and individuals. Activity is closely related to cardiovascular health. This study aimed to investigate the relationship between the varying domains of PA, like occupation-related Physical Activity (OPA), transportation-related Physical Activity (TPA), leisure-time Physical Activity (LTPA), and Sedentary Activity (SA) with stroke. Methods Our analysis included 30,400 participants aged 20+ years from 2007 to 2018 National Health and Nutrition Examination Survey (NHANES). Stroke was identified based on the participant's self-reported diagnoses from previous medical consultations, and PA and SA were self-reported. Multivariable logistic and restricted cubic spline models were used to assess the associations. Results Participants achieving PA guidelines (performing PA more than 150 min/week) were 35.7% less likely to have a stroke based on both the total PA (odds ratio [OR] 0.643, 95% confidence interval [CI] 0.523-0.790) and LTPA (OR 0.643, 95% CI 0.514-0.805), while OPA or TPA did not demonstrate lower stroke risk. Furthermore, participants with less than 7.5 h/day SA levels were 21.6% (OR 0.784, 95% CI 0.665-0.925) less likely to have a stroke. The intensities of total PA and LTPA exhibited nonlinear U-shaped associations with stroke risk. In contrast, those of OPA and TPA showed negative linear associations, while SA intensities were positively linearly correlated with stroke risk. Conclusions LTPA, but not OPA or TPA, was associated with a lower risk of stroke at any amount, suggesting that significant cardiovascular health would benefit from increased PA. Additionally, the positive association between SA and stroke indicated that prolonged sitting was detrimental to cardiovascular health. Overall, increased PA within a reasonable range reduces the risk of stroke, while increased SA elevates it.

physics.med-ph↗

arXiv2023

CoGANPPIS: A Coevolution-enhanced Global Attention Neural Network for Protein-Protein Interaction Site Prediction

Protein-protein interactions are of great importance in biochemical processes. Accurate prediction of protein-protein interaction sites (PPIs) is crucial for our understanding of biological mechanism. Although numerous approaches have been developed recently and achieved gratifying results, there are still two limitations: (1) Most existing models have excavated a number of useful input features, but failed to take coevolutionary features into account, which could provide clues for inter-residue relationships; (2) The attention-based models only allocate attention weights for neighboring residues, instead of doing it globally, which may limit the model's prediction performance since some residues being far away from the target residues might also matter. We propose a coevolution-enhanced global attention neural network, a sequence-based deep learning model for PPIs prediction, called CoGANPPIS. Specifically, CoGANPPIS utilizes three layers in parallel for feature extraction: (1) Local-level representation aggregation layer, which aggregates the neighboring residues' features as the local feature representation; (2) Global-level representation learning layer, which employs a novel coevolution-enhanced global attention mechanism to allocate attention weights to all residues on the same protein sequences; (3) Coevolutionary information learning layer, which applies CNN & pooling to coevolutionary information to obtain the coevolutionary profile representation. Then, the three outputs are concatenated and passed into several fully connected layers for the final prediction. Extensive experiments on two benchmark datasets have been conducted, demonstrating that our proposed model achieves the state-of-the-art performance.

q-bio.QM↗

arXiv2020

Forecasting and evaluating intervention of Covid-19 in the World

When the Covid-19 pandemic enters dangerous new phase, whether and when to take aggressive public health interventions to slow down the spread of COVID-19. To develop the artificial intelligence (AI) inspired methods for real-time forecasting and evaluating intervention strategies to curb the spread of Covid-19 in the World. A modified auto-encoder for modeling the transmission dynamics of the epidemics is developed and applied to the surveillance data of cumulative and new Covid-19 cases and deaths from WHO, as of March 16, 2020. The average errors of 5-step forecasting were 2.5%. The total peak number of cumulative cases and new cases, and the maximum number of cumulative cases in the world with later intervention (comprehensive public health intervention is implemented 4 weeks later) could reach 75,249,909, 10,086,085, and 255,392,154, respectively. The case ending time was January 10, 2021. However, the total peak number of cumulative cases and new cases and the maximum number of cumulative cases in the world with one week later intervention were reduced to 951,799, 108,853 and 1,530,276, respectively. Duration time of the Covid-19 spread would be reduced from 356 days to 232 days. The case ending time was September 8, 2020. We observed that delaying intervention for one month caused the maximum number of cumulative cases to increase 166.89 times, and the number of deaths increase from 53,560 to 8,938,725. We will face disastrous consequences if immediate action to intervene is not taken.

q-bio.PE↗

arXiv2020

Artificial Intelligence Forecasting of Covid-19 in China

BACKGROUND An alternative to epidemiological models for transmission dynamics of Covid-19 in China, we propose the artificial intelligence (AI)-inspired methods for real-time forecasting of Covid-19 to estimate the size, lengths and ending time of Covid-19 across China. METHODS We developed a modified stacked auto-encoder for modeling the transmission dynamics of the epidemics. We applied this model to real-time forecasting the confirmed cases of Covid-19 across China. The data were collected from January 11 to February 27, 2020 by WHO. We used the latent variables in the auto-encoder and clustering algorithms to group the provinces/cities for investigating the transmission structure. RESULTS We forecasted curves of cumulative confirmed cases of Covid-19 across China from Jan 20, 2020 to April 20, 2020. Using the multiple-step forecasting, the estimated average errors of 6-step, 7-step, 8-step, 9-step and 10-step forecasting were 1.64%, 2.27%, 2.14%, 2.08%, 0.73%, respectively. We predicted that the time points of the provinces/cities entering the plateau of the forecasted transmission dynamic curves varied, ranging from Jan 21 to April 19, 2020. The 34 provinces/cities were grouped into 9 clusters. CONCLUSIONS The accuracy of the AI-based methods for forecasting the trajectory of Covid-19 was high. We predicted that the epidemics of Covid-19 will be over by the middle of April. If the data are reliable and there are no second transmissions, we can accurately forecast the transmission dynamics of the Covid-19 across the provinces/cities in China. The AI-inspired methods are a powerful tool for helping public health planning and policymaking.

q-bio.OT↗

arXiv2019

Shared Causal Paths underlying Alzheimer's dementia and Type 2 Diabetes

Background: Although Alzheimer's disease (AD) is a central nervous system disease and type 2 diabetes mellitus (T2DM) is a metabolic disorder, an increasing number of genetic epidemiological studies show clear link between AD and T2DM. The current approach to uncovering the shared pathways between AD and T2DM involves association analysis; however, such analyses lack power to discover the mechanisms of the diseases. Methods: We develop novel statistical methods to shift the current paradigm of genetic analysis from association analysis to deep causal inference for uncovering the shared mechanisms between AD and T2DM, and develop pipelines to infer multilevel omics causal networks which lead to shifting the current paradigm of genetic analysis from genetic analysis alone to integrated causal genomic, epigenomic, transcriptional and phenotypic data analysis. To discover common causal paths from genetic variants to AD and T2DM, we also develop algorithms that can automatically search the causal paths from genetic variants to diseases and Results: The proposed methods and algorithms are applied to ROSMAP dataset with 432 individuals who simultaneously had genotype, RNA-seq, DNA methylation and some phenotypes. We construct multi-omics causal networks and identify 13 shared causal genes, 16 shared causal pathways between AD and T2DM, and 754 gene expression and 101 gene methylation nodes that were connected to both AD and T2DM in multi-omics causal networks. Conclusions: The results of application of the proposed pipelines for identifying causal paths to real data analysis of AD and T2DM provided strong evidence to support the link between AD and T2DM and unraveled causal mechanism to explain this link.

q-bio.MN↗

arXiv2018

Bivariate Causal Discovery and its Applications to Gene Expression and Imaging Data Analysis

The mainstream of research in genetics, epigenetics and imaging data analysis focuses on statistical association or exploring statistical dependence between variables. Despite their significant progresses in genetic research, understanding the etiology and mechanism of complex phenotypes remains elusive. Using association analysis as a major analytical platform for the complex data analysis is a key issue that hampers the theoretic development of genomic science and its application in practice. Causal inference is an essential component for the discovery of mechanical relationships among complex phenotypes. Many researchers suggest making the transition from association to causation. Despite its fundamental role in science, engineering and biomedicine, the traditional methods for causal inference require at least three variables. However, quantitative genetic analysis such as QTL, eQTL, mQTL, and genomic-imaging data analysis requires exploring the causal relationships between two variables. This paper will focus on bivariate causal discovery. We will introduce independence of cause and mechanism (ICM) as a basic principle for causal inference, algorithmic information theory and additive noise model (ANM) as major tools for bivariate causal discovery. Large-scale simulations will be performed to evaluate the feasibility of the ANM for bivariate causal discovery. To further evaluate their performance for causal inference, the ANM will be applied to the construction of gene regulatory networks. Also, the ANM will be applied to trait-imaging data analysis to illustrate three scenarios: presence of both causation and association, presence of association while absence of causation, and presence of causation, while lack of association between two variables.

q-bio.GN↗