arXiv subjects

Junde Wu

Publications and source records attributed to Junde Wu.

At least 55 records · Page 3Linked to original sources

arXiv2025

SurgVLM: A Large Vision-Language Model and Systematic Evaluation Benchmark for Surgical Intelligence

Foundation models have achieved transformative success across biomedical domains by enabling holistic understanding of multimodal data. However, their application in surgery remains underexplored. Surgical intelligence presents unique challenges - requiring surgical visual perception, temporal analysis, and reasoning. Existing general-purpose vision-language models fail to address these needs due to insufficient domain-specific supervision and the lack of a large-scale high-quality surgical database. To bridge this gap, we propose SurgVLM, one of the first large vision-language foundation models for surgical intelligence, where this single universal model can tackle versatile surgical tasks. To enable this, we construct a large-scale multimodal surgical database, SurgVLM-DB, comprising over 1.81 million frames with 7.79 million conversations, spanning more than 16 surgical types and 18 anatomical structures. We unify and reorganize 23 public datasets across 10 surgical tasks, followed by standardizing labels and doing hierarchical vision-language alignment to facilitate comprehensive coverage of gradually finer-grained surgical tasks, from visual perception, temporal analysis, to high-level reasoning. Building upon this comprehensive dataset, we propose SurgVLM, which is built upon Qwen2.5-VL, and undergoes instruction tuning to 10+ surgical tasks. We further construct a surgical multimodal benchmark, SurgVLM-Bench, for method evaluation. SurgVLM-Bench consists of 6 popular and widely-used datasets in surgical domain, covering several crucial downstream tasks. Based on SurgVLM-Bench, we evaluate the performance of our SurgVLM (3 SurgVLM variants: SurgVLM-7B, SurgVLM-32B, and SurgVLM-72B), and conduct comprehensive comparisons with 14 mainstream commercial VLMs (e.g., GPT-4o, Gemini 2.0 Flash, Qwen2.5-Max).

cs.CV↗

arXiv2025

Quantum algorithms for calculating determinant and inverse of matrix and solving linear algebraic systems

We propose quantum algorithms, purely quantum in nature, for calculating the determinant and inverse of an $(N-1)\times (N-1)$ matrix (depth is $O(N^2\log N)$) which is a simple modification of the algorithm for calculating the determinant of an $N\times N$ matrix (depth is $O(N\log^2 N)$. The basic idea is to encode each row of the matrix into a pure state of some quantum system. In addition, we use the representation of the elements of the inverse matrix in terms of algebraic complements. This algorithm together with that for matrix multiplication { proposed earlier} yields the algorithm for solving systems of linear algebraic equations (depth is $O(N\log^2 N)$. Measurement of the ancilla state with output 1 (probability is $\sim 2^{-O(N\log N)}$) removes the garbage acquired during calculation. Appropriate circuits for all three algorithms are presented and have the same estimation $O(N\log N)$ for the space (number of qubits in the circuit).

quant-ph↗

arXiv2025

ReSurgSAM2: Referring Segment Anything in Surgical Video via Credible Long-term Tracking

Surgical scene segmentation is critical in computer-assisted surgery and is vital for enhancing surgical quality and patient outcomes. Recently, referring surgical segmentation is emerging, given its advantage of providing surgeons with an interactive experience to segment the target object. However, existing methods are limited by low efficiency and short-term tracking, hindering their applicability in complex real-world surgical scenarios. In this paper, we introduce ReSurgSAM2, a two-stage surgical referring segmentation framework that leverages Segment Anything Model 2 to perform text-referred target detection, followed by tracking with reliable initial frame identification and diversity-driven long-term memory. For the detection stage, we propose a cross-modal spatial-temporal Mamba to generate precise detection and segmentation results. Based on these results, our credible initial frame selection strategy identifies the reliable frame for the subsequent tracking. Upon selecting the initial frame, our method transitions to the tracking stage, where it incorporates a diversity-driven memory mechanism that maintains a credible and diverse memory bank, ensuring consistent long-term tracking. Extensive experiments demonstrate that ReSurgSAM2 achieves substantial improvements in accuracy and efficiency compared to existing methods, operating in real-time at 61.2 FPS. Our code and datasets will be available at https://github.com/jinlab-imvr/ReSurgSAM2.

cs.CV↗

arXiv2025

Arbitrary state creation via controlled measurement

The initial state creation is a starting point of many quantum algorithms and usually is considered as a separate subroutine not included into the algorithm itself. There are many algorithms aimed on creation of special class of states. Our algorithm allows creating an arbitrary $n$-qubit pure quantum superposition state with precision of $m$-decimals (binary representation) for each probability amplitude. The algorithm uses one-qubit rotations, Hadamard transformations and C-NOT operations with multi-qubit controls. However, the crucial operation is the final controlled measurement of the ancilla state that removes the garbage part of the superposition state and allows to avoid the problem of small success probability in that measurement. We emphasize that rotation angles are predicted in advance by the required precision and therefore there is no classical calculation supplementing quantum algorithm. The depth and space of the algorithm growth with $n$ as, respectively, $O(2^n n)$ and $O(n)$. This algorithm can be a subroutine generating the required input state in various algorithms, in particular, in matrix-manipulation algorithms developed earlier.

quant-ph↗

arXiv2025

MedVLM-R1: Incentivizing Medical Reasoning Capability of Vision-Language Models (VLMs) via Reinforcement Learning

Reasoning is a critical frontier for advancing medical image analysis, where transparency and trustworthiness play a central role in both clinician trust and regulatory approval. Although Medical Visual Language Models (VLMs) show promise for radiological tasks, most existing VLMs merely produce final answers without revealing the underlying reasoning. To address this gap, we introduce MedVLM-R1, a medical VLM that explicitly generates natural language reasoning to enhance transparency and trustworthiness. Instead of relying on supervised fine-tuning (SFT), which often suffers from overfitting to training distributions and fails to foster genuine reasoning, MedVLM-R1 employs a reinforcement learning framework that incentivizes the model to discover human-interpretable reasoning paths without using any reasoning references. Despite limited training data (600 visual question answering samples) and model parameters (2B), MedVLM-R1 boosts accuracy from 55.11% to 78.22% across MRI, CT, and X-ray benchmarks, outperforming larger models trained on over a million samples. It also demonstrates robust domain generalization under out-of-distribution tasks. By unifying medical image analysis with explicit reasoning, MedVLM-R1 marks a pivotal step toward trustworthy and interpretable AI in clinical practice. Inference model is available at: https://huggingface.co/JZPeterPan/MedVLM-R1.

cs.CV↗

arXiv2025

Surgical SAM 2: Real-time Segment Anything in Surgical Video by Efficient Frame Pruning

Surgical video segmentation is a critical task in computer-assisted surgery and is vital for enhancing surgical quality and patient outcomes. Recently, the Segment Anything Model 2 (SAM2) framework has shown superior advancements in image and video segmentation. However, SAM2 struggles with efficiency due to the high computational demands of processing high-resolution images and complex and long-range temporal dynamics in surgical videos. To address these challenges, we introduce Surgical SAM 2 (SurgSAM2), an advanced model to utilize SAM2 with an Efficient Frame Pruning (EFP) mechanism, to facilitate real-time surgical video segmentation. The EFP mechanism dynamically manages the memory bank by selectively retaining only the most informative frames, reducing memory usage and computational cost while maintaining high segmentation accuracy. Our extensive experiments demonstrate that SurgSAM2 significantly improves both efficiency and segmentation accuracy compared to the vanilla SAM2. Remarkably, SurgSAM2 achieves a 3$\times$ FPS compared with SAM2, while also delivering state-of-the-art performance after fine-tuning with lower-resolution data. These advancements establish SurgSAM2 as a leading model for surgical video analysis, making real-time surgical video segmentation in resource-constrained environments a reality. Our source code is available at https://github.com/jinlab-imvr/Surgical-SAM-2.

cs.CV↗

arXiv2025

Implementation of Quantum Fourier Transform and Quantum Hashing for a Quantum Device with Arbitrary Qubits Connection Graphs

In the paper, we consider quantum circuits for Quantum fingerprinting (quantum hashing) and quantum Fourier transform (QFT) algorithms. Quantum fingerprinting (quantum hashing) is a well-known technique for comparing large objects using small images. The QFT algorithm is a very popular technique used in many algorithms. We present a generic method for constructing quantum circuits for these algorithms for quantum devices with restrictions. Many quantum devices (for example, based on superconductors) have restrictions on applying two-qubit gates. The restrictions are presented by a qubits connection graph. Typically, researchers consider only the linear nearest neighbor (LNN) architecture, but current devices have more complex graphs. We present a method for arbitrary connected graphs that minimizes the number of CNOT gates in the circuit. The heuristic version of the method is fast enough and works with $O(n^5)$ time complexity, where $n$ is the number of qubits. The certain version of the algorithm has an exponential time complexity that is $O(n^22^n)$. We compare quantum circuits built by our algorithm with quantum circuits optimized for specific graphs that are Linear-nearest-neighbor (LNN) architecture, ``sun'' (a cycle with tails, presented by 16-qubit IBMQ device) and ``two joint suns'' (two joint cycles with tails, presented by 27-qubit IBMQ device). Our generic method gives similar results with little bit more CNOT gates. At the same time, our method allows us to construct a circuit for arbitrary connected graphs.

quant-ph↗

arXiv2025

Remarks on controlled measurement and quantum algorithm for calculating Hermitian conjugate

We present two new aspects for the recently proposed algorithms for matrix manipulating based on the special encoding the matrix elements into the superposition state of a quantum system. First aspect is the controlled measurement which allows to avoid the problem of small access probability to the required ancilla state at the final step of algorithms needed to remove the garbage of the states. Application of controlled measurement to the earlier developed algorithm is demonstrated. The second aspect is the algorithm for calculating the Hermitian conjugate of an arbitrary matrix, which supplements the algorithms proposed earlier. The appropriate circuits are presented.

quant-ph↗

arXiv2025

Analysis of a nonlinear free boundary problem modeling the radial growth of two-layer tumors

In this paper we study a nonlinear free boundary problem on the radial growth of a two-layer solid tumor with a quiescent core. The tumor surface and its inner interface separating the proliferating cells and the quiescent cells are both free boundaries. By deeply analyzing their relationship and employing the maximum principle, we show this problem is globally well-posed and prove the existence of a unique positive threshold $σ^*$ such that the problem admits a unique stationary solution with a quiescent core if and only if the externally supplied nutrient $\barσ> σ^*$. The stationary solution is globally asymptotically stable. The formation of the quiescent core and its interesting connection with the necrotic core are also given.

math.AP↗

arXiv2024

Relative Quantum Resource Theory and Operational Applications in Subchannel Discrimination

A central problem in quantum resource theory is to give operational meaning to quantum resources that can provide clear advantages in certain physical tasks compared to the convex set of resource-free states. We propose to extend this basic principle by defining the relative superiority of resources over a specific convex set of resource states, also provide a relative advantage in physical tasks based on this extended principle. This allows the generalized robustness measure to quantify the relative maximal advantage due to a given resource state over a specific convex set of resource states in the subchannel discrimination, thereby showing that the operational interpretation of resource measures also holds in a relative perspective. In addition, we offer a new framework for defining the deficiency of a given state in physical tasks compared to the set of maximum resource states. The geometric measure we provide satisfies the conditions of the framework for quantum coherence and entanglement, and it accurately quantifies the minimal disadvantage due to a given state compared to maximumresourcestates inthe subchannel discrimination in certain situations. These two extensions and new interpretations expand the scope of quantum resource theories and provide a more comprehensive operational interpretation.

quant-ph↗

arXiv2024

Medical SAM 2: Segment medical images as video via Segment Anything Model 2

Medical image segmentation plays a pivotal role in clinical diagnostics and treatment planning, yet existing models often face challenges in generalization and in handling both 2D and 3D data uniformly. In this paper, we introduce Medical SAM 2 (MedSAM-2), a generalized auto-tracking model for universal 2D and 3D medical image segmentation. The core concept is to leverage the Segment Anything Model 2 (SAM2) pipeline to treat all 2D and 3D medical segmentation tasks as a video object tracking problem. To put it into practice, we propose a novel \emph{self-sorting memory bank} mechanism that dynamically selects informative embeddings based on confidence and dissimilarity, regardless of temporal order. This mechanism not only significantly improves performance in 3D medical image segmentation but also unlocks a \emph{One-Prompt Segmentation} capability for 2D images, allowing segmentation across multiple images from a single prompt without temporal relationships. We evaluated MedSAM-2 on five 2D tasks and nine 3D tasks, including white blood cells, optic cups, retinal vessels, mandibles, coronary arteries, kidney tumors, liver tumors, breast cancer, nasopharynx cancer, vestibular schwannoma, mediastinal lymph nodules, cerebral artery, inferior alveolar nerve, and abdominal organs, comparing it against state-of-the-art (SOTA) models in task-tailored, general and interactive segmentation settings. Our findings demonstrate that MedSAM-2 surpasses a wide range of existing models and updates new SOTA on several benchmarks. The code is released on the project page: https://supermedintel.github.io/Medical-SAM2/.

cs.CV↗

arXiv2024

Matrix manipulations via unitary transformations and ancilla-state measurements

We propose protocols for calculating inner product, matrix addition and matrix multiplication based on multiqubit Toffoli-type and the simplest one-qubit operations and employ ancilla measurements to remove all garbage of calculations. The depth (runtime) of the addition protocol is $O(1)$ and that of other protocols logarithmically increases with the dimensionality of the considered matrices.

quant-ph↗

arXiv2024

Medical Graph RAG: Towards Safe Medical Large Language Model via Graph Retrieval-Augmented Generation

We introduce a novel graph-based Retrieval-Augmented Generation (RAG) framework specifically designed for the medical domain, called \textbf{MedGraphRAG}, aimed at enhancing Large Language Model (LLM) capabilities for generating evidence-based medical responses, thereby improving safety and reliability when handling private medical data. Graph-based RAG (GraphRAG) leverages LLMs to organize RAG data into graphs, showing strong potential for gaining holistic insights from long-form documents. However, its standard implementation is overly complex for general use and lacks the ability to generate evidence-based responses, limiting its effectiveness in the medical field. To extend the capabilities of GraphRAG to the medical domain, we propose unique Triple Graph Construction and U-Retrieval techniques over it. In our graph construction, we create a triple-linked structure that connects user documents to credible medical sources and controlled vocabularies. In the retrieval process, we propose U-Retrieval which combines Top-down Precise Retrieval with Bottom-up Response Refinement to balance global context awareness with precise indexing. These effort enable both source information retrieval and comprehensive response generation. Our approach is validated on 9 medical Q\&A benchmarks, 2 health fact-checking benchmarks, and one collected dataset testing long-form generation. The results show that MedGraphRAG consistently outperforms state-of-the-art models across all benchmarks, while also ensuring that responses include credible source documentation and definitions. Our code is released at: https://github.com/MedicineToken/Medical-Graph-RAG.

cs.CV↗

arXiv2024

MedUHIP: Towards Human-In-the-Loop Medical Segmentation

Although segmenting natural images has shown impressive performance, these techniques cannot be directly applied to medical image segmentation. Medical image segmentation is particularly complicated by inherent uncertainties. For instance, the ambiguous boundaries of tissues can lead to diverse but plausible annotations from different clinicians. These uncertainties cause significant discrepancies in clinical interpretations and impact subsequent medical interventions. Therefore, achieving quantitative segmentations from uncertain medical images becomes crucial in clinical practice. To address this, we propose a novel approach that integrates an \textbf{uncertainty-aware model} with \textbf{human-in-the-loop interaction}. The uncertainty-aware model proposes several plausible segmentations to address the uncertainties inherent in medical images, while the human-in-the-loop interaction iteratively modifies the segmentation under clinician supervision. This collaborative model ensures that segmentation is not solely dependent on automated techniques but is also refined through clinician expertise. As a result, our approach represents a significant advancement in the field which enhances the safety of medical image segmentation. It not only offers a comprehensive solution to produce quantitative segmentation from inherent uncertain medical images, but also establishes a synergistic balance between algorithmic precision and clincian knowledge. We evaluated our method on various publicly available multi-clinician annotated datasets: REFUGE2, LIDC-IDRI and QUBIQ. Our method showcases superior segmentation capabilities, outperforming a wide range of deterministic and uncertainty-aware models. We also demonstrated that our model produced significantly better results with fewer interactions compared to previous interactive models. We will release the code to foster further research in this area.

eess.IV↗

arXiv2024

MGI: Multimodal Contrastive pre-training of Genomic and Medical Imaging

Medicine is inherently a multimodal discipline. Medical images can reflect the pathological changes of cancer and tumors, while the expression of specific genes can influence their morphological characteristics. However, most deep learning models employed for these medical tasks are unimodal, making predictions using either image data or genomic data exclusively. In this paper, we propose a multimodal pre-training framework that jointly incorporates genomics and medical images for downstream tasks. To address the issues of high computational complexity and difficulty in capturing long-range dependencies in genes sequence modeling with MLP or Transformer architectures, we utilize Mamba to model these long genomic sequences. We aligns medical images and genes using a self-supervised contrastive learning approach which combines the Mamba as a genetic encoder and the Vision Transformer (ViT) as a medical image encoder. We pre-trained on the TCGA dataset using paired gene expression data and imaging data, and fine-tuned it for downstream tumor segmentation tasks. The results show that our model outperformed a wide range of related methods.

cs.CV↗

arXiv2024

One-Prompt to Segment All Medical Images

Large foundation models, known for their strong zero-shot generalization, have excelled in visual and language applications. However, applying them to medical image segmentation, a domain with diverse imaging types and target labels, remains an open challenge. Current approaches, such as adapting interactive segmentation models like Segment Anything Model (SAM), require user prompts for each sample during inference. Alternatively, transfer learning methods like few/one-shot models demand labeled samples, leading to high costs. This paper introduces a new paradigm toward the universal medical image segmentation, termed 'One-Prompt Segmentation.' One-Prompt Segmentation combines the strengths of one-shot and interactive methods. In the inference stage, with just \textbf{one prompted sample}, it can adeptly handle the unseen task in a single forward pass. We train One-Prompt Model on 64 open-source medical datasets, accompanied by the collection of over 3,000 clinician-labeled prompts. Tested on 14 previously unseen datasets, the One-Prompt Model showcases superior zero-shot segmentation capabilities, outperforming a wide range of related methods. The code and data is released as https://github.com/KidsWithTokens/one-prompt.

eess.IV↗

arXiv2024

Not just Birds and Cars: Generic, Scalable and Explainable Models for Professional Visual Recognition

Some visual recognition tasks are more challenging then the general ones as they require professional categories of images. The previous efforts, like fine-grained vision classification, primarily introduced models tailored to specific tasks, like identifying bird species or car brands with limited scalability and generalizability. This paper aims to design a scalable and explainable model to solve Professional Visual Recognition tasks from a generic standpoint. We introduce a biologically-inspired structure named Pro-NeXt and reveal that Pro-NeXt exhibits substantial generalizability across diverse professional fields such as fashion, medicine, and art-areas previously considered disparate. Our basic-sized Pro-NeXt-B surpasses all preceding task-specific models across 12 distinct datasets within 5 diverse domains. Furthermore, we find its good scaling property that scaling up Pro-NeXt in depth and width with increasing GFlops can consistently enhances its accuracy. Beyond scalability and adaptability, the intermediate features of Pro-NeXt achieve reliable object detection and segmentation performance without extra training, highlighting its solid explainability. We will release the code to foster further research in this area.

cs.CV↗

arXiv2023

Quantum Algorithms for the Shortest Common Superstring and Text Assembling Problems

In this paper, we consider two versions of the Text Assembling problem. We are given a sequence of strings $s^1,\dots,s^n$ of total length $L$ that is a dictionary, and a string $t$ of length $m$ that is texts. The first version of the problem is assembling $t$ from the dictionary. The second version is the ``Shortest Superstring Problem''(SSP) or the ``Shortest Common Superstring Problem''(SCS). In this case, $t$ is not given, and we should construct the shortest string (we call it superstring) that contains each string from the given sequence as a substring. These problems are connected with the sequence assembly method for reconstructing a long DNA sequence from small fragments. For both problems, we suggest new quantum algorithms that work better than their classical counterparts. In the first case, we present a quantum algorithm with $O(m+\log m\sqrt{nL})$ running time. In the case of SSP, we present a quantum algorithm with running time $O(n^3 1.728^n +L +\sqrt{L}n^{1.5}+\sqrt{L}n\log^2L\log^2n)$.

quant-ph↗