arXiv subjects

Changchun Yang

Publications and source records attributed to Changchun Yang.

At least 19 recordsLinked to original sources

arXiv2026

TRACE: An Evidence-Grounded Benchmark for Safety Evaluation of Large Reasoning Models

Large Reasoning Models (LRMs) generate intermediate reasoning traces that may contain unsafe content, even when their final responses appear safe. Guardrail models are designed to detect and block unsafe content, yet existing benchmarks for unsafe content detection focus primarily on prompts and final responses, leaving reasoning traces largely unexamined. Moreover, these benchmarks typically provide only binary safety labels, without evidence annotations that justify the judgments. To address these limitations, we introduce TRACE, an evidence-grounded safety evaluation benchmark that covers the entire LRM inference pipeline: prompts, reasoning traces, and final responses. TRACE includes prompts in two languages spanning nine risk categories and ten attack strategies. For each prompt, four LRMs generate reasoning traces and final responses, and we annotate the safety of each component and extract supporting evidence from the corresponding source text. Evaluating 18 guardrail models on TRACE reveals that safety judgment for reasoning traces is substantially more challenging than for prompts or final responses, and that current models struggle to accurately extract supporting evidence. These findings highlight the need for guardrail models that can reliably detect and precisely localize unsafe content across the LRM inference pipeline.

cs.AI↗

arXiv2026

Perturb-and-Restore: Simulation-driven Structural Augmentation Framework for Imbalance Chromosomal Anomaly Detection

Detecting structural chromosomal abnormalities is crucial for accurate diagnosis and management of genetic disorders. However, collecting sufficient structural abnormality data is extremely challenging and costly in clinical practice, and not all abnormal types can be readily collected. As a result, deep learning approaches face significant performance degradation due to the severe imbalance and scarcity of abnormal chromosome data. To address this challenge, we propose a Perturb-and-Restore (P&R), a simulation-driven structural augmentation framework that effectively alleviates data imbalance in chromosome anomaly detection. The P&R framework comprises two key components: (1) Structure Perturbation and Restoration Simulation, which generates synthetic abnormal chromosomes by perturbing chromosomal banding patterns of normal chromosomes followed by a restoration diffusion network that reconstructs continuous chromosome content and edges, thus eliminating reliance on rare abnormal samples; and (2) Energy-guided Adaptive Sampling, an energy score-based online selection strategy that dynamically prioritizes high-quality synthetic samples by referencing the energy distribution of real samples. To evaluate our method, we construct a comprehensive structural anomaly dataset consisting of over 260,000 chromosome images, including 4,242 abnormal samples spanning 24 categories. Experimental results demonstrate that the P&R framework achieves state-of-the-art (SOTA) performance, surpassing existing methods with an average improvement of 8.92% in sensitivity, 8.89% in precision, and 13.79% in F1-score across all categories.

cs.CV↗

arXiv2026

Colon-Bench: An Agentic Workflow for Scalable Dense Lesion Annotation in Full-Procedure Colonoscopy Videos

Early screening via colonoscopy is critical for colon cancer prevention, yet developing robust AI systems for this domain is hindered by the lack of densely annotated, long-sequence video datasets. Existing datasets predominantly focus on single-class polyp detection and lack the rich spatial, temporal, and linguistic annotations required to evaluate modern Multimodal Large Language Models (MLLMs). To address this critical gap, we introduce Colon-Bench, generated via a novel multi-stage agentic workflow. Our pipeline seamlessly integrates temporal proposals, bounding-box tracking, AI-driven visual confirmation, and human-in-the-loop review to scalably annotate full-procedure videos. The resulting verified benchmark is unprecedented in scope, encompassing 528 videos, 14 distinct lesion categories (including polyps, ulcers, and bleeding), over 300,000 bounding boxes, 213,000 segmentation masks, and 133,000 words of clinical descriptions. We utilize Colon-Bench to rigorously evaluate state-of-the-art MLLMs across lesion classification, Open-Vocabulary Video Object Segmentation (OV-VOS), and video Visual Question Answering (VQA). The MLLM results demonstrate surprisingly high localization performance in medical domains compared to SAM-3. Finally, we analyze common VQA errors from MLLMs to introduce a novel "colon-skill" prompting strategy, improving zero-shot MLLM performance by up to 9.7% across most MLLMs. The dataset and the code are available at https://abdullahamdi.com/colon-bench .

eess.IV↗

arXiv2025

Structural Prognostic Event Modeling for Multimodal Cancer Survival Analysis

The integration of histology images and gene profiles has shown great promise for improving survival prediction in cancer. However, current approaches often struggle to model intra- and inter-modal interactions efficiently and effectively due to the high dimensionality and complexity of the inputs. A major challenge is capturing critical prognostic events that, though few, underlie the complexity of the observed inputs and largely determine patient outcomes. These events, manifested as high-level structural signals such as spatial histologic patterns or pathway co-activations, are typically sparse, patient-specific, and unannotated, making them inherently difficult to uncover. To address this, we propose SlotSPE, a slot-based framework for structural prognostic event modeling. Specifically, inspired by the principle of factorial coding, we compress each patient's multimodal inputs into compact, modality-specific sets of mutually distinctive slots using slot attention. By leveraging these slot representations as encodings for prognostic events, our framework enables both efficient and effective modeling of complex intra- and inter-modal interactions, while also facilitating seamless incorporation of biological priors that enhance prognostic relevance. Extensive experiments on ten cancer benchmarks show that SlotSPE outperforms existing methods in 8 out of 10 cohorts, achieving an overall improvement of 2.9%. It remains robust under missing genomic data and delivers markedly improved interpretability through structured event decomposition.

cs.CV↗

arXiv2025

PAST: A multimodal single-cell foundation model for histopathology and spatial transcriptomics in cancer

While pathology foundation models have transformed cancer image analysis, they often lack integration with molecular data at single-cell resolution, limiting their utility for precision oncology. Here, we present PAST, a pan-cancer single-cell foundation model trained on 20 million paired histopathology images and single-cell transcriptomes spanning multiple tumor types and tissue contexts. By jointly encoding cellular morphology and gene expression, PAST learns unified cross-modal representations that capture both spatial and molecular heterogeneity at the cellular level. This approach enables accurate prediction of single-cell gene expression, virtual molecular staining, and multimodal survival analysis directly from routine pathology slides. Across diverse cancers and downstream tasks, PAST consistently exceeds the performance of existing approaches, demonstrating robust generalizability and scalability. Our work establishes a new paradigm for pathology foundation models, providing a versatile tool for high-resolution spatial omics, mechanistic discovery, and precision cancer research.

q-bio.QM↗

arXiv2025

An Inclusive Foundation Model for Generalizable Cytogenetics in Precision Oncology

Chromosome analysis is vital for diagnosing genetic disorders and guiding cancer therapy decisions through the identification of somatic clonal aberrations. However, developing an AI model are hindered by the overwhelming complexity and diversity of chromosomal abnormalities, requiring extensive annotation efforts, while automated methods remain task-specific and lack generalizability due to the scarcity of comprehensive datasets spanning diverse resource conditions. Here, we introduce CHROMA, a foundation model for cytogenomics, designed to overcome these challenges by learning generalizable representations of chromosomal abnormalities. Pre-trained on over 84,000 specimens (~4 million chromosomal images) via self-supervised learning, CHROMA outperforms other methods across all types of abnormalities, even when trained on fewer labelled data and more imbalanced datasets. By facilitating comprehensive mapping of instability and clonal leisons across various aberration types, CHROMA offers a scalable and generalizable solution for reliable and automated clinical analysis, reducing the annotation workload for experts and advancing precision oncology through the early detection of rare genomic abnormalities, enabling broad clinical AI applications and making advanced genomic analysis more accessible.

q-bio.QM↗

arXiv2024

Improving Representation of High-frequency Components for Medical Visual Foundation Models

Foundation models have recently attracted significant attention for their impressive generalizability across diverse downstream tasks. However, these models are demonstrated to exhibit great limitations in representing high-frequency components and fine-grained details. In many medical imaging tasks, the precise representation of such information is crucial due to the inherently intricate anatomical structures, sub-visual features, and complex boundaries involved. Consequently, the limited representation of prevalent foundation models can result in significant performance degradation or even failure in these tasks. To address these challenges, we propose a novel pretraining strategy, named Frequency-advanced Representation Autoencoder (Frepa). Through high-frequency masking and low-frequency perturbation combined with adversarial learning, Frepa encourages the encoder to effectively represent and preserve high-frequency components in the image embeddings. Additionally, we introduce an innovative histogram-equalized image masking strategy, extending the Masked Autoencoder approach beyond ViT to other architectures such as Swin Transformer and convolutional networks. We develop Frepa across nine medical modalities and validate it on 32 downstream tasks for both 2D images and 3D volume data. Without fine-tuning, Frepa can outperform other self-supervised pretraining methods and, in some cases, even surpasses task-specific trained models. This improvement is particularly significant for tasks involving fine-grained details, such as achieving up to a +15% increase in DSC for retina vessel segmentation and a +7% increase in IoU for lung nodule detection. Further experiments quantitatively reveal that Frepa enables superior high-frequency representations and preservation in the embeddings, underscoring its potential for developing more generalized and universal medical image foundation models.

eess.IV↗

arXiv2024

Deep learning-driven pulmonary artery and vein segmentation reveals demography-associated vasculature anatomical differences

Pulmonary artery-vein segmentation is crucial for disease diagnosis and surgical planning and is traditionally achieved by Computed Tomography Pulmonary Angiography (CTPA). However, concerns regarding adverse health effects from contrast agents used in CTPA have constrained its clinical utility. In contrast, identifying arteries and veins using non-contrast CT, a conventional and low-cost clinical examination routine, has long been considered impossible. Here we propose a High-abundant Pulmonary Artery-vein Segmentation (HiPaS) framework achieving accurate artery-vein segmentation on both non-contrast CT and CTPA across various spatial resolutions. HiPaS first performs spatial normalization on raw CT volumes via a super-resolution module, and then iteratively achieves segmentation results at different branch levels by utilizing the lower-level vessel segmentation as a prior for higher-level vessel segmentation. We trained and validated HiPaS on our established multi-centric dataset comprising 1,073 CT volumes with meticulous manual annotations. Both quantitative experiments and clinical evaluation demonstrated the superior performance of HiPaS, achieving an average dice score of 91.8% and a sensitivity of 98.0%. Further experiments showed the non-inferiority of HiPaS segmentation on non-contrast CT compared to segmentation on CTPA. Employing HiPaS, we have conducted an anatomical study of pulmonary vasculature on 11,784 participants in China (six sites), discovering a new association of pulmonary vessel anatomy with sex, age, and disease states: vessel abundance suggests a significantly higher association with females than males with slightly decreasing with age, and is also influenced by certain diseases, under the controlling of lung volumes.

cs.CV↗

arXiv2022

Efficient Bayesian Uncertainty Estimation for nnU-Net

The self-configuring nnU-Net has achieved leading performance in a large range of medical image segmentation challenges. It is widely considered as the model of choice and a strong baseline for medical image segmentation. However, despite its extraordinary performance, nnU-Net does not supply a measure of uncertainty to indicate its possible failure. This can be problematic for large-scale image segmentation applications, where data are heterogeneous and nnU-Net may fail without notice. In this work, we introduce a novel method to estimate nnU-Net uncertainty for medical image segmentation. We propose a highly effective scheme for posterior sampling of weight space for Bayesian uncertainty estimation. Different from previous baseline methods such as Monte Carlo Dropout and mean-field Bayesian Neural Networks, our proposed method does not require a variational architecture and keeps the original nnU-Net architecture intact, thereby preserving its excellent performance and ease of use. Additionally, we boost the segmentation performance over the original nnU-Net via marginalizing multi-modal posterior models. We applied our method on the public ACDC and M&M datasets of cardiac MRI and demonstrated improved uncertainty estimation over a range of baseline methods. The proposed method further strengthens nnU-Net for medical image segmentation in terms of both segmentation accuracy and quality control.

cs.CV↗

arXiv2022

Photoacoustic Digital Skin: Generation and Simulation of Human Skin Vascular for Quantitative Image Analysis

Photoacoustic computed tomography (PACT) is a hybrid imaging modality, which combines the high optical contrast of pure optical imaging and the high penetration depth of ultrasound imaging. However, photoacoustic image dataset with good quality and large quantity is lacking. In this paper, we mainly talk about how to generate a practical photoacoustic dataset. Firstly, we extracted 389 3D vessel volumes from CT whole-lung scan database, and enhanced the blood vessel structures. Then for each 3D vessel volume, we embedded it into a three-layer cubic phantom to formulate a skin tissue model, which includes epidermis, dermis, and hypodermis. The vessel volume was placed randomly in dermis layer in 10 different ways. Thus, 3890 3D skin tissue phantoms were generated. Then we assigned optical properties for the four kinds of tissue types. Monte-Carlo optical simulations were deployed to obtain the optical fluence distribution. Then acoustic propagation simulations were deployed to obtain the photoacoustic initial pressure. Universal back-projection algorithm was used to reconstruct the photoacoustic images. This dataset could be used for deep learning-based photoacoustic image reconstruction, classification, registration, quantitative image analysis.

physics.med-ph↗

arXiv2021

Photoacoustic digital brain: numerical modelling and image reconstruction via deep learning

Photoacoustic tomography (PAT) is a newly developed medical imaging modality, which combines the advantages of pure optical imaging and ultrasound imaging, owning both high optical contrast and deep penetration depth. Very recently, PAT is studied in human brain imaging. Nevertheless, while ultrasound waves are passing through the human skull tissues, the strong acoustic attenuation and aberration will happen, which causes photoacoustic signals' distortion. In this work, we use 10 magnetic resonance angiography (MRA) human brain volumes, and manually segment them to obtain the 2D human brain numerical phantoms for PAT. The numerical phantoms contain six kinds of tissues which are scalp, skull, white matter, gray matter, blood vessel and cerebral cortex. For every numerical phantom, optical properties are assigned to every kind of tissues. Then, Monte-Carlo based optical simulation is deployed to obtain the photoacoustic initial pressure. Then, we made two k-wave simulation cases: one takes inhomogeneous medium and uneven sound velocity into consideration, and the other not. Then we use the sensor data of the former one as the input of U-net, and the sensor data of the latter one as the output of U-net to train the network. We randomly choose 7 human brain PA sinograms as the training dataset and 3 human brain PA sinograms as the testing set. The testing result shows that our method could correct the skull acoustic aberration and obtain the blood vessel distribution inside the human brain satisfactorily.

physics.med-ph↗

arXiv2021

AS-Net: Fast Photoacoustic Reconstruction with Multi-feature Fusion from Sparse Data

Photoacoustic (PA) imaging is a biomedical imaging modality capable of acquiring high-contrast images of optical absorption at depths much greater than traditional optical imaging techniques. However, practical instrumentation and geometry limit the number of available acoustic sensors surrounding the imaging target, which results in the sparsity of sensor data. Conventional PA image reconstruction methods give severe artifacts when they are applied directly to the sparse PA data. In this paper, we firstly propose to employ a novel signal processing method to make sparse PA raw data more suitable for the neural network, concurrently speeding up image reconstruction. Then we propose Attention Steered Network (AS-Net) for PA reconstruction with multi-feature fusion. AS-Net is validated on different datasets, including simulated photoacoustic data from fundus vasculature phantoms and experimental data from in vivo fish and mice. Notably, the method is also able to eliminate some artifacts present in the ground truth for in vivo data. Results demonstrated that our method provides superior reconstructions at a faster speed.

cs.CV↗

arXiv2021

Compressed Sensing for Photoacoustic Computed Tomography Using an Untrained Neural Network

Photoacoustic (PA) computed tomography (PACT) shows great potentials in various preclinical and clinical applications. A great number of measurements are the premise that obtains a high-quality image, which implies a low imaging rate or a high system cost. The artifacts or sidelobes could pollute the image if we decrease the number of measured channels or limit the detected view. In this paper, a novel compressed sensing method for PACT using an untrained neural network is proposed, which decreases half number of the measured channels and recoveries enough details. This method uses a neural network to reconstruct without the requirement for any additional learning based on the deep image prior. The model can reconstruct the image only using a few detections with gradient descent. Our method can cooperate with other existing regularization, and further improve the quality. In addition, we introduce a shape prior to easily converge the model to the image. We verify the feasibility of untrained network based compressed sensing in PA image reconstruction, and compare this method with a conventional method using total variation minimization. The experimental results show that our proposed method outperforms 32.72% (SSIM) with the traditional compressed sensing method in the same regularization. It could dramatically reduce the requirement for the number of transducers, by sparsely sampling the raw PA data, and improve the quality of PA image significantly.

cs.CV↗

arXiv2021

A Jointed Feature Fusion Framework for Photoacoustic Reconstruction

Photoacoustic (PA) computed tomography (PACT) reconstructs the initial pressure distribution from raw PA signals. The standard reconstruction of medical image could cause the artifacts due to interferences or ill-posed setup. Recently, deep learning has been used to reconstruct the PA image with ill-posed conditions. Most works remove the artifacts from image domain, and compensate the limited-view from dataset. In this paper, we propose a jointed feature fusion framework (JEFF-Net) based on deep learning to reconstruct the PA image using limited-view data. The cross-domain features from limited-view position-wise data and the reconstructed image are fused by a backtracked supervision. Specifically, our results could generate superior performance, whose artifacts are drastically reduced in the output compared to ground-truth (full-view reconstructed result). In this paper, a quarter position-wise data (32 channels) is fed into model, which outputs another 3-quarters-view data (96 channels). Moreover, two novel losses are designed to restrain the artifacts by sufficiently manipulating superposed data. The numerical and in-vivo results have demonstrated the superior performance of our method to reconstruct the full-view image without artifacts. Finally, quantitative evaluations show that our proposed method outperformed the ground-truth in some metrics.

cs.CV↗

arXiv2020

Superposed Wave (s-Wave): Accelerating Photoacoustic Simulation

Photoacoustic imaging develops very fast in recent years due to its superior performance in many preclinical and clinical applications. However, it is still in a developing stage, and a lot of experiments have to be performed in a simulation setting. To simulate photoacoustic imaging in a computer, k-Wave is currently the most popular MATLAB toolbox. Lots of research groups choose k-Wave toolbox to perform the forward projection process, which also can be described as forward model. However, by solving complex partial differential equation, k-Wave suffers a lot from computation time. To accelerate photoacoustic simulation, in this paper, we propose a straightforward simulation approach based on superposed Wave (s-Wave). Specifically, we treat the initial pressure distribution as a set of single pixels. Then by pre-obtaining a standard sensor data from single pixel, we can easily use loop and multiplication operators to change phase and amplitude of sensor data for given pixels. We use three different 2D samples and two 3D samples to test the time cost. The result of our proposed s-Wave method shows much less time consumption compared with k-wave. Especially in a sparse configuration in 3D, s-Wave is more than 2000 times faster than k-Wave, whiling getting nearly same sensor data.

physics.med-ph↗

arXiv2020

Deep learning for photoacoustic imaging: a survey

Machine learning has been developed dramatically and witnessed a lot of applications in various fields over the past few years. This boom originated in 2009, when a new model emerged, that is, the deep artificial neural network, which began to surpass other established mature models on some important benchmarks. Later, it was widely used in academia and industry. Ranging from image analysis to natural language processing, it fully exerted its magic and now become the state-of-the-art machine learning models. Deep neural networks have great potential in medical imaging technology, medical data analysis, medical diagnosis and other healthcare issues, and is promoted in both pre-clinical and even clinical stages. In this review, we performed an overview of some new developments and challenges in the application of machine learning to medical image analysis, with a special focus on deep learning in photoacoustic imaging. The aim of this review is threefold: (i) introducing deep learning with some important basics, (ii) reviewing recent works that apply deep learning in the entire ecological chain of photoacoustic imaging, from image reconstruction to disease diagnosis, (iii) providing some open source materials and other resources for researchers interested in applying deep learning to photoacoustic imaging.

cs.CV↗

arXiv2020

Limited-view Photoacoustic Imaging Reconstruction With Dual Domain Inputs Under Mutual Information Constraint

Based on photoacoustic effect, photoacoustic tomography is developing very fast in recent years, and becoming an important imaging tool for both preclinical and clinical studies. With enough ultrasound transducers placed around the biological tissue, PAT can provide both deep penetration and high image contrast by hybrid usage of light and sound. However, considering space and measurement environmental limitations, transducers are always placed in a limited-angle way, which means that the other side without transducer coverage suffers severe information loss. With conventional image reconstruction algorithms, the limited-view tissue induces artifacts and information loss, which may cause doctors misdiagnosis or missed diagnosis. In order to solve limited-view PA imaging reconstruction problem, we propose to use both time domain and frequency domain reconstruction algorithms to get delay-and-sum (DAS) image inputs and k-space image inputs. These dual domain images share nearly same texture information but different artifact information, which can teach network how to distinguish these two kinds of information at input level. In this paper, we propose Dual Domain Unet (DuDoUnet) with specially designed Information Sharing Block (ISB), which can further share two domains' information and distinguish artifacts. Besides, we use mutual information (MI) with an auxiliary network, whose inputs and outputs are both ground truth, to compensate prior knowledge of limited-view PA inputs. The proposed method is verified with a public clinical database, and shows superior results with SSIM = 93.5622% and PSNR = 20.8859.

eess.IV↗

arXiv2020

Deep Learning Enables Robust and Precise Light Focusing on Treatment Needs

If light passes through the body tissues, focusing only on areas where treatment needs, such as tumors, will revolutionize many biomedical imaging and therapy technologies. So how to focus light through deep inhomogeneous tissues overcoming scattering is Holy Grail in biomedical areas. In this paper, we use deep learning to learn and accelerate the process of phase pre-compensation using wavefront shaping. We present an approach (LoftGAN, light only focuses on treatment needs) for learning the relationship between phase domain X and speckle domain Y . Our goal is not just to learn an inverse mapping F:Y->X such that we can know the corresponding X needed for imaging Y like most work, but also to make focusing that is susceptible to disturbances more robust and precise by ensuring that the phase obtained can be forward mapped back to speckle. So we introduce different constraints to enforce F(Y)=X and H(F(Y))=Y with the transmission mapping H:X->Y. Both simulation and physical experiments are performed to investigate the effects of light focusing to demonstrate the effectiveness of our method and comparative experiments prove the crucial improvement of robustness and precision. Codes are available at https://github.com/ChangchunYang/LoftGAN.

eess.IV↗