arXiv subjects

Na Su

Publications and source records attributed to Na Su.

15 recordsLinked to original sources

arXiv2026

Propagating Structural Guidance: Synthesizing Fluorescein Angiography from Fundus Images and Sparse OCT Scans

Fundus fluorescein angiography (FFA) is critical for assessing retinal vascular abnormalities, but its acquisition is invasive and not always feasible. In contrast, color fundus photography (CFP) is non-invasive and widely accessible, which has motivated studies on CFP-to-FFA synthesis. However, prior works rely solely on CFP surface texture, fundamentally limiting the ability to reconstruct functional vascular information and subtle pathological changes. To address this, we propose a novel framework that synthesizes FFA from CFP with structural guidance provided by optical coherence tomography (OCT). We construct a multi-modal retinal imaging dataset with paired CFP, FFA, and OCT from 3,676 patient eyes--the first tri-modally aligned dataset in retinal imaging. To bridge the spatial gap between OCT and fundus modalities, we propose a Spatially Aligned Cross-Modal Fusion (SACMF) module that projects depth-resolved OCT features onto the fundus plane and injects them into the CFP encoder via adaptive layer normalization. Beyond feature fusion, we further introduce Token-wise Cross-Modality Alignment (TCMA), a token-level contrastive learning strategy that explicitly aligns CFP and FFA representations at corresponding spatial positions. Our method achieves superior synthesis performance compared to state-of-the-art methods. Moreover, extensive experiments demonstrate that the FFA images synthesized by our approach bring greater improvements in downstream disease diagnosis performance than existing methods, highlighting the clinical potential of our approach as a non-invasive decision-support tool in routine workflows. The code is available at https://github.com/while-plus/OCT-guide-FFA-Syn.

cs.CV↗

arXiv2025

Bridging the Gap in Ophthalmic AI: MM-Retinal-Reason Dataset and OphthaReason Model toward Dynamic Multimodal Reasoning

Multimodal large language models (MLLMs) have recently demonstrated remarkable reasoning abilities with reinforcement learning paradigm. Although several multimodal reasoning models have been explored in the medical domain, most of them focus exclusively on basic reasoning, which refers to shallow inference based on visual feature matching. However, real-world clinical diagnosis extends beyond basic reasoning, demanding reasoning processes that integrate heterogeneous clinical information (such as chief complaints and medical history) with multimodal medical imaging data. To bridge this gap, we introduce MM-Retinal-Reason, the first ophthalmic multimodal dataset with the full spectrum of perception and reasoning. It encompasses both basic reasoning tasks and complex reasoning tasks, aiming to enhance visual-centric fundamental reasoning capabilities and emulate realistic clinical thinking patterns. Building upon MM-Retinal-Reason, we propose OphthaReason, the first ophthalmology-specific multimodal reasoning model with step-by-step reasoning traces. To enable flexible adaptation to both basic and complex reasoning tasks, we specifically design a novel method called Uncertainty-Aware Dynamic Thinking (UADT), which estimates sample-level uncertainty via entropy and dynamically modulates the model's exploration depth using a shaped advantage mechanism. Comprehensive experiments demonstrate that our model achieves state-of-the-art performance on both basic and complex reasoning tasks, outperforming general-purpose MLLMs, medical MLLMs, RL-based medical MLLMs, and ophthalmic MLLMs by at least 24.92\%, 15.00\%, 21.20\%, and 17.66\%. Project Page: \href{https://github.com/lxirich/OphthaReason}{link}.

cs.AI↗

arXiv2025

VISTA: Enhancing Vision-Text Alignment in MLLMs via Cross-Modal Mutual Information Maximization

Current multimodal large language models (MLLMs) face a critical challenge in modality alignment, often exhibiting a bias towards textual information at the expense of other modalities like vision. This paper conducts a systematic information-theoretic analysis of the widely used cross-entropy loss in MLLMs, uncovering its implicit alignment objective. Our theoretical investigation reveals that this implicit objective has inherent limitations, leading to a degradation of cross-modal alignment as text sequence length increases, thereby hindering effective multimodal information fusion. To overcome these drawbacks, we propose Vision-Text Alignment (VISTA), a novel approach guided by our theoretical insights. VISTA introduces an explicit alignment objective designed to maximize cross-modal mutual information, preventing the degradation of visual alignment. Notably, VISTA enhances the visual understanding capabilities of existing MLLMs without requiring any additional trainable modules or extra training data, making it both efficient and practical. Our method significantly outperforms baseline models across more than a dozen benchmark datasets, including VQAv2, MMStar, and MME, paving the way for new directions in MLLM modal alignment research.

cs.CV↗

arXiv2025

Tuning Magnetism of Metal-Organic Framework by Different Types of Pressure

The properties of metal-organic frameworks (MOFs) are expected to be sensitive to external pressures because of their inherently flexible structures. Although pressure-driven structural transitions have been intensively studied, the influence of pressure on magnetism has been less exploited for MOFs. Especially, the efficiency of applied pressure may strongly depend on the pressure-transmitting medium (PTM) which may have a complex interaction with MOFs. Here, we report the distinctive effects of different types of pressure, including isotropic hydrostatic pressure, quasi-hydrostatic pressure, and uniaxial pressure, on the anisotropic magnetism of the perovskite MOF, [CH3NH3][Co(COOH)3]. It is found that the hydrostatic pressure has the minimal effect whereas the uniaxial pressure has the highest efficiency in tuning magnetization and magnetic anisotropy of the MOF. First-principles calculations reveal that the applied low pressures (less than 11 kbar) do not induce notable lattice distortion in the framework or the superexchange path. Instead, the modulation of hydrogen bonds is identified as a critical factor for pressure tuning of magnetization and anisotropy. These findings underscore the potential applications of directional pressure in precisely controlling the magnetic and electronic properties of MOFs.

cond-mat.mtrl-sci↗

arXiv2025

From Visuals to Vocabulary: Establishing Equivalence Between Image and Text Token Through Autoregressive Pre-training in MLLMs

While MLLMs perform well on perceptual tasks, they lack precise multimodal alignment, limiting performance. To address this challenge, we propose Vision Dynamic Embedding-Guided Pretraining (VDEP), a hybrid autoregressive training paradigm for MLLMs. Utilizing dynamic embeddings from the MLP following the visual encoder, this approach supervises image hidden states and integrates image tokens into autoregressive training. Existing MLLMs primarily focused on recovering information from textual inputs, often neglecting the effective processing of image data. In contrast, the key improvement of this work is the reinterpretation of multimodal alignment as a process of recovering information from input data, with particular emphasis on reconstructing detailed visual features.The proposed method seamlessly integrates into standard models without architectural changes. Experiments on 13 benchmarks show VDEP outperforms baselines, surpassing existing methods.

cs.CV↗

arXiv2025

MM-Retinal V2: Transfer an Elite Knowledge Spark into Fundus Vision-Language Pretraining

Vision-language pretraining (VLP) has been investigated to generalize across diverse downstream tasks for fundus image analysis. Although recent methods showcase promising achievements, they significantly rely on large-scale private image-text data but pay less attention to the pretraining manner, which limits their further advancements. In this work, we introduce MM-Retinal V2, a high-quality image-text paired dataset comprising CFP, FFA, and OCT image modalities. Then, we propose a novel fundus vision-language pretraining model, namely KeepFIT V2, which is pretrained by integrating knowledge from the elite data spark into categorical public datasets. Specifically, a preliminary textual pretraining is adopted to equip the text encoder with primarily ophthalmic textual knowledge. Moreover, a hybrid image-text knowledge injection module is designed for knowledge transfer, which is essentially based on a combination of global semantic concepts from contrastive learning and local appearance details from generative learning. Extensive experiments across zero-shot, few-shot, and linear probing settings highlight the generalization and transferability of KeepFIT V2, delivering performance competitive to state-of-the-art fundus VLP models trained on large-scale private image-text datasets. Our dataset and model are publicly available via https://github.com/lxirich/MM-Retinal.

cs.CV↗

arXiv2024

Memory-efficient High-resolution OCT Volume Synthesis with Cascaded Amortized Latent Diffusion Models

Optical coherence tomography (OCT) image analysis plays an important role in the field of ophthalmology. Current successful analysis models rely on available large datasets, which can be challenging to be obtained for certain tasks. The use of deep generative models to create realistic data emerges as a promising approach. However, due to limitations in hardware resources, it is still difficulty to synthesize high-resolution OCT volumes. In this paper, we introduce a cascaded amortized latent diffusion model (CA-LDM) that can synthesis high-resolution OCT volumes in a memory-efficient way. First, we propose non-holistic autoencoders to efficiently build a bidirectional mapping between high-resolution volume space and low-resolution latent space. In tandem with autoencoders, we propose cascaded diffusion processes to synthesize high-resolution OCT volumes with a global-to-local refinement process, amortizing the memory and computational demands. Experiments on a public high-resolution OCT dataset show that our synthetic data have realistic high-resolution and global features, surpassing the capabilities of existing methods. Moreover, performance gains on two down-stream fine-grained segmentation tasks demonstrate the benefit of the proposed method in training deep learning models for medical imaging tasks. The code is public available at: https://github.com/nicetomeetu21/CA-LDM.

eess.IV↗

arXiv2024

Pressure tuning of hydrogen bond ordering in the metal-organic framework [(CH3)2NH2]Mn(HCOO)3

The influence of pressure on the hydrogen bond ordering in the perovskite metal-organic framework [(CH3)2NH2]Mn(HCOO)3 has been investigated by dielectric, pyroelectric adn magnetic measurements in a piston-cylinder cell. Under ambient pressure the ordering of hydrogen bonds takes place at TC = 188 K and induces a first-order ferroelectric phase transition. With increasing pressure to p = 3.92 kbar, the order-disorder transition shifts to a lower temperature and retains the first-order ferroelectric nature. However, under higher pressures, the ordering process of hydrogen bonds is split into two transitions: a broad antiferroelectric transition at high temperature and a first-order ferroelectric transition at low temperature. With increasing pressure, the antiferroelectric phase is enhanced whereas the ferroelectric phase is greatly suppressed, which implies that compression of the perovskite framework favors antiparallel arrangement of the hydrogen bonds. The canted anti-ferromagnetic transition was almost unchanged when pressure up to 10.85 kbar. Our study demonstrated that the perovskite metal-organic frameworks are more sensitive to external pressure than conventional perovskite oxides so that their electric properties can be easily tuned by pressure.

cond-mat.mtrl-sci↗

arXiv2024

Resonant Quantum Magnetodielectric Effect in Multiferroic Metal-Organic Framework [CH3NH3]Co(HCOO)3

We report the observation of both resonant quantum tunneling of magnetization (RQTM) and resonant quantum magnetodielectric (RQMD) effect in the perovskite multiferroic metal-organic framework [CH3NH3]Co(HCOO)3. An intrinsic magnetic phase separation emerges at low temperatures due to hydrogen-bond-modified long range super-exchange interaction, leading to the coexistence of canted antiferromagnetic order and single-ion magnet. Subsequently, a stair-shaped magnetic hysteresis loop along the [101] direction characterizing the RQTM appears below the magnetic blocking temperature. More interestingly, the magnetic field dependence of dielectric permittivity exhibits pronounced negative peaks at the critical fields corresponding to the RQTM, a phenomenon termed the RQMD effect which enables electrical detection of the RQTM. These intriguing properties make the multiferroic metal-organic framework a promising candidate for solid-state quantum computing.

cond-mat.mtrl-sci↗

arXiv2024

Pressure tunable magnetic skyrmion phase in Co8Zn8Mn4 single crystals

In a magnetic skyrmion phase, magnetic moments form vortex-like topological textures which are of both fundamental and industrial interests. In $\beta$-Mn-type Co-Zn-Mn alloys, chrial magnetic skyrmions emerge above room temperature, providing a unique system for studying the skrymion physics and exploring spintronics applications. However, the magnetic skyrmion phase is typically confined in a narrow and limited temperature ($T$) and magnetic field ($H$) range. Here, we demonstrate that hydrostatic pressure can expand the skyrmion phase in the $T-H$ phase diagram of single-crystalline Co$_8$Zn$_8$Mn$_4$. At ambient pressure, signatures of skyrmions are seen within $T\sim302-308$ K and $H\sim50-100$ Oe. Applying a moderate pressure of 6 kbar extends this range to $T\sim300-310$ K and $H\sim50-150$ Oe. However, further escalation of pressure to 10 kbar results in a slight contraction of the skyrmion phase. These findings underscore the sensitivity of the skyrmion phase in Co$_8$Zn$_8$Mn$_4$ to external pressures, and hint at the potential of strain engineering, particularly in $\beta$-Mn-type Co-Zn-Mn thin films, as a promising avenue to customize the skyrmion phase.

cond-mat.mtrl-sci↗

arXiv2023

Adjustable Robust Transformer for High Myopia Screening in Optical Coherence Tomography

Myopia is a manifestation of visual impairment caused by an excessively elongated eyeball. Image data is critical material for studying high myopia and pathological myopia. Measurements of spherical equivalent and axial length are the gold standards for identifying high myopia, but the available image data for matching them is scarce. In addition, the criteria for defining high myopia vary from study to study, and therefore the inclusion of samples in automated screening efforts requires an appropriate assessment of interpretability. In this work, we propose a model called adjustable robust transformer (ARTran) for high myopia screening of optical coherence tomography (OCT) data. Based on vision transformer, we propose anisotropic patch embedding (APE) to capture more discriminative features of high myopia. To make the model effective under variable screening conditions, we propose an adjustable class embedding (ACE) to replace the fixed class token, which changes the output to adapt to different conditions. Considering the confusion of the data at high myopia and low myopia threshold, we introduce the label noise learning strategy and propose a shifted subspace transition matrix (SST) to enhance the robustness of the model. Besides, combining the two structures proposed above, the model can provide evidence for uncertainty evaluation. The experimental results demonstrate the effectiveness and reliability of the proposed method. Code is available at: https://github.com/maxiao0234/ARTran.

cs.CV↗

arXiv2022

Tuning of Quantum Paraelectricity of M-type Hexaferrite BaFe12O19 by External Parameters

M-type hexaferrite BaFe12O19 was recently reported to be a new type of quantum paraelectrics with triangular lattice by showing a low temperature dielectric plateau due to quantum fluctuation. It has also been proposed to have a possible quantum-dipole liquid ground state. To suppress its quantum fluctuations and reach a possible quantum critical point, we have tuned its quantum paraelectricity in three ways: (i) 57Fe isotope replacement; (ii) in-plane compressive strain; and (iii) hydrostatic pressure. It is found that 95% 57Fe replacement and the in-plane strain are more effective to drive its ground state closer to a critical region by inducing a peak feature in the temperature dependence of dielectric constant. In contrast, the application of hydrostatic pressure pushed the system away from the quantum critical point by gradually suppressing the plateau feature in dielectric constant. Our combined efforts reveal the potential of the M-type hexaferrites for studying the quantum critical behaviors.

cond-mat.str-el↗

arXiv2021

Probing magnetic symmetry in antiferromagnetic Fe4Nb2O9 single crystals by linear magnetoelectric tensor

In the present study, we investigated magnetodielectric, magnetoelectric (ME), and angular-dependent polarization in single-crystal Fe4Nb2O9. The magnetodielectric effects in epsilon(x) (x//[100]), epsilon(y) (y//[120]), and epsilon(z) (z//[001]) are found to be significant only around T-N approximate to 95 K when magnetic fields are applied along three orthogonal x-, y- (y//[120]), and z directions. The finite polarization P-x, P-y, and P-z of 70, 100, and 30 mu C/m(2), respectively, can be induced in the antiferromagnetic phase when a finite magnetic field up to 9 T is applied along the three orthogonal directions. The angular-dependent polarization measurements verify the dominating linear ME effects below T-N. From the above experimental results, a linear ME tensor a(ij) with all nine nonzero components can be inferred, demonstrating a much lower magnetic point group of -1' for the canted antiferromagnetic configuration.

cond-mat.mtrl-sci↗

arXiv2021

Reentrance of spin-driven ferroelectricity through rotational tunneling of ammonium

Quantum effects fundamentally engender exotic physical phenomena in macroscopic systems, which advance next-generation technological applications. Rotational tunneling that represents the quantum phenomenon of the librational motion of molecules is ubiquitous in hydrogen-contained materials. However, its direct manifestation in realizing macroscopic physical properties is elusive. Here we report an observation of reentrant ferroelectricity under low pressure that is mediated by the rotational tunneling of ammonium ions in molecule-based (NH$_4$)$_2$FeCl$_5 \cdot$H$_2$O. Applying a small pressure leads to a transition from spin-driven ferroelectricity to paraelectricity coinciding with the stabilization of a collinear magnetic phase. Such a transition is attributed to the hydrogen bond fluctuations via the rotational tunneling of ammonium groups as supported by theoretical calculations. Higher pressure lifts the quantum fluctuations and leads to a reentrant ferroelectric phase concomitant with another incommensurate magnetic phase. These results demonstrate that the rotational tunneling emerges as a new route to control magnetic-related properties in soft magnets, opening avenues for designing multi-functional materials and realizing potential quantum control.

cond-mat.mtrl-sci↗

arXiv2016

Criticality-Enhanced Magnetocaloric Effect in Quantum Spin Chain Material Copper Nitrate

Low-dimensional quantum magnets, due to the existence of abundant exotic quantum phases therein and experimental feasibilities in laboratories, continues intriguing people in condensed matter physics. In this work, a comprehensive study of Cu(NO$_3$)$_2$ $\cdot$ 2.5H$_2$O (copper nitrate hemipentahydrate, CN), a spin chain material, is performed with multi-technique approach including thermal tensor network (TTN) simulations, first-principles calculations, as well as magnetization measurements in experiments. Employing a cutting-edge TTN method developed in the present work, we determine the couplings $J=5.13$ K, $\alpha=0.23(1)$ and Land\'e factors $g_{\parallel}=2.31$, $g_{\perp}=2.14$ in an alternating Heisenberg antiferromagnetic chain model, with which one can fit strikingly well the magnetothermodynamic properties. Part of the fitted experimental data are measured on the single-crystal CN specimens synthesized by us. Based on first-principles calculations, we reveal explicitly the spin chain scenario in CN by displaying the calculated electron density distributions, from which the distinct superexchange paths are visualized. On top of that, we investigated the magnetocaloric effect (MCE) in CN by calculating its isentropes and magnetic Gr\"ueisen parameter (GP). Prominent quantum-criticality-enhanced MCE was uncovered, the TTN simulations are in good agreements with measured isentropic lines in the sub-Kelvin region. We propose that CN is potentially a very promising quantum critical coolant, due to the remarkably enhanced MCE near both critical fields of moderate strengths as 2.87 and 4.08 T, respectively.

cond-mat.str-el↗