IEEE transactions on medical imaging最新文献

{"title":"Frequency-Aware Causal Regularization for Multiple Instance Learning in Whole Slide Image Classification","authors":"Dawei Fan;Lifang Wei;Mingyue Han;Tao Xu;Xuemei Qiu;Yanping Chen;Changcai Yang;Riqing Chen","doi":"10.1109/TMI.2026.3697015","DOIUrl":"10.1109/TMI.2026.3697015","url":null,"abstract":"Whole slide image (WSI) classification is a critical task in computational pathology and is aimed at providing automated diagnostic support through high-resolution tissue image analysis. In weakly supervised WSI classification scenarios, the main challenge concerns the traditional multiple instance learning (MIL) methods, which rely on instance-level embeddings aggregated by an attention-based pooling mechanism. These methods often depend on data-driven statistical correlations, leading to misalignments between their attention allocation schemes and histopathological diagnostic regions and reducing the resulting prediction reliability. To address this, we propose frequency-aware causal regularized multiple instance learning (FC-MIL), an innovative framework combining that combines frequency-aware attention (FAA) and causal regularization (CR). FAA extracts more granular, fine-grained histological textures by jointly modeling spatial- and frequency- domain features, whereas CR introduces feature-level counterfactual perturbations as an intervention-inspired regularizer in the latent space, encouraging the model to rely less on spurious correlations and more on invariant pathological cues. Experimental results obtained on four WSI datasets show that FC-MIL outperforms the state-of-the-art MIL methods in terms of both accuracy and interpretability. Our source code is available at <uri>https://github.com/7FFDW/FCMIL</uri>","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4287-4302"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148038747","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Investigation of Drug Responses in 3-D Tumor Spheroid Models Using Two-Photon Scanning Structured Illumination Super-Resolution Microscopy With Frequency-Specific Denoising Enhancement","authors":"Meiting Wang;Xinran Li;Peng Du;Yuye Wang;Jiajie Chen;Ying Wu;Ying Long;Bingchun Jiang;Junle Qu;Bruce Zhi Gao;Xiao Peng;Yonghong Shao","doi":"10.1109/TMI.2026.3698950","DOIUrl":"10.1109/TMI.2026.3698950","url":null,"abstract":"Two-dimensional cell culture models have long been a cornerstone of biomedical research; however, they often fail to accurately replicate the in vivo environment. In recent years, three-dimensional (3D) cell cultures, particularly 3D spheroid models, have gained recognition for their ability to better mimic the complexities of the in vivo environment, making them valuable tools for studying cellular behavior and responses. Tumor spheroids, in particular, have significant applications in anticancer therapy evaluation, providing a more physiologically relevant model by simulating the spatial architecture and microenvironment of tumors. However, due to the limitations imposed by optical diffraction and background noise in 3D imaging, traditional imaging methods are unable to accurately resolve the growth, morphological changes, and drug responses of tumor spheroids. To address this issue, super-resolution imaging technologies have emerged. Structured illumination microscopy (SIM) combined with reconstruction algorithms can effectively enhance resolution, but challenges such as limited light penetration of single-photon imaging and high background noise remain in 3D imaging. In this paper, an advanced SIM technology with large depth and low noise 3D imaging capability is developed. This study introduces a novel frequency-specific denoising method (FSDM) to effectively reduce noise through adjusting the weights of high-frequency signals to preserve image details. The FSDM optimization significantly reduces background interference from deeper tissue layers, improving image details and the overall quality of 3D imaging. For the first time, scanning SIM is integrated with two-photon microscopy (TPEF-SIM) for 3D imaging, leveraging the strengths of both techniques to enhance resolution and overcome light penetration limitations.","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4395-4405"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148152581","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing Brain Signal Generation Through a Hybrid Approach Integrating Reinforcement Learning and Diffusion Models","authors":"Yang An;Yuhao Tong;Weikai Wang;Steven W. Su","doi":"10.1109/TMI.2026.3696676","DOIUrl":"10.1109/TMI.2026.3696676","url":null,"abstract":"Developing a reliable EEG-based Brain Computer Interface (BCI) system typically requires large and diverse training datasets, but collecting sufficient data remains challenging due to subject fatigue and inter-individual variability. To address these limitations, this study proposes a reinforcement learning-enhanced EEG diffusion (RLED) framework for adaptive data augmentation in endogenous EEG tasks, with a focus on motor imagery and emotion recognition. The framework integrates a reinforcement learning mechanism to dynamically regulate the diffusion training process and achieve a flexible balance among temporal, spectral, and class-related features. Experiments on four datasets demonstrate that the proposed method generates high-quality synthetic EEG signals and consistently improves classification performance. These findings show that the proposed RLED framework may serve as a promising tool for EEG data augmentation and generalization in practical BCI applications.","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4254-4276"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148019697","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Self-Supervised T2WI-Bridged Framework for Liver Segmentation and PDFF Prediction From US Images","authors":"Dong Zhang;Qi Zeng;Septimiu E. Salcudean;Z. Jane Wang","doi":"10.1109/TMI.2026.3698415","DOIUrl":"10.1109/TMI.2026.3698415","url":null,"abstract":"Proton Density Fat Fraction (PDFF) is the gold standard for non-invasive fatty liver diagnosis, but its reliance on Magnetic Resonance Imaging (MRI) limits broad clinical applicability. Motivated by the accessibility of B-mode Ultrasound (US) in fatty liver assessment, we propose a novel framework for liver segmentation and PDFF prediction from US images. To enhance generalization ability despite limited paired US-PDFF data, our framework integrates a cross-task self-supervised pretext task that extracts semantic features to guide echo intensity capture, benefiting both liver segmentation and PDFF prediction. To address the noise and artifacts inherent in US images, our framework leverages T2-weighted imaging (T2WI) exclusively during training to establish a feature bridge between US and PDFF, thereby enhancing PDFF prediction. Once trained, the model relies solely on US for inference, making it a practical and cost-effective alternative to MRI-based PDFF estimation. Additionally, our framework introduces an uncertainty-augmented adversarial loss function to refine liver boundary delineation, further improving segmentation and PDFF prediction accuracy. Experimental results demonstrate that our method outperforms state-of-the-art methods in liver segmentation and PDFF prediction; and in a specific application study, our predicted PDFF achieves accuracy comparable to real PDFF for hepatic steatosis classification, highlighting its clinical potential. The full source code and detailed documentation are publicly available at <uri>https://github.com/D0ngZhang/SSTB</uri>","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4351-4365"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148059140","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DGPDT: Detection-Guided and Prompt-Driven Transformers for Automated and Generalizable Cobb Angle Estimation","authors":"Chih-Yi Lu;Chen-Hung Tu;Chun-Yi Hsieh;Chi-Kuang Feng;I-Yun Lisa Hsieh","doi":"10.1109/TMI.2026.3701624","DOIUrl":"10.1109/TMI.2026.3701624","url":null,"abstract":"Accurate Cobb angle measurement is essential for scoliosis assessment but remains labor-intensive and observer-dependent. We introduce DGPDT (Detection-Guided Prompt-driven Transformer), a unified transformer-based framework that integrates vertebra detection and foundation-model segmentation for generalizable spinal analysis. A Roboflow Detection Transformer (RF-DETR) with a DINOv2 backbone localizes vertebrae, followed by post-processing to ensure anatomical continuity. The resulting bounding boxes serve as automatic prompts for a fine-tuned Segment Anything Model 2.1 (SAM 2.1), which generates high-resolution vertebral masks. Cobb angles are then computed from vertebral masks, enabling estimation of both main and compensatory curves. Evaluations on the in-house (TVGH-SpineXR) and external (SpineWeb-16) datasets demonstrate encouraging performance on both internal and external datasets, achieving mean Dice coefficients of 0.944 and 0.781, respectively, and mean absolute Cobb angle errors of approximately 2–3° in-domain and 4.93° under cross-domain evaluation. Despite being trained solely on TVGH-SpineXR, DGPDT maintains accuracy comparable to models trained directly on the benchmark dataset. By coupling detection-guided prompting with transformer-based segmentation, DGPDT achieves a clinically acceptable mean absolute error (<inline-formula> <tex-math>$lt 5^{circ }text {)}$ </tex-math></inline-formula>, suggesting good reproducibility and potential applicability beyond the training dataset.","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4444-4455"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148213942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"UPGRADE-Net: Unsupervised Sinogram-Domain Data-Consistent Network for Metal Artifact Reduction","authors":"Zhan Wu;Yikun Zhang;Yongjie Guo;Hui Tang;Yinsheng Li;Huazhong Shu;Yan Xi;Yi Zhang;Gouenou Coatrieux;Yang Chen","doi":"10.1109/TMI.2025.3630832","DOIUrl":"10.1109/TMI.2025.3630832","url":null,"abstract":"Computed tomography (CT) scanners are widely used to obtain detailed internal images in clinical diagnosis. Highly attenuated metallic implants resulting from strong and energy-dependent attenuation cause metal artifacts in CT scanning. However, current supervised deep network-based metal artifact reduction (MAR) methods hardly generalize in clinical diagnosis and treatment because of difficult acquisition for the paired artifact-affected and artifact-free data. In addition, these deep model-based methods cannot ensure the sinogram-domain data consistency for the exact metal trace inpainting. To address the above problems, we propose an UnsuPervised sinoGRam-domAin Data-consistEnt network for MAR, <italic>i.e.</i>, UPGRADE-Net. First, UPGRADE-Net fully leverages the prior knowledge to guide the generative conditional diffusion model for fine-grained metal trace inpainting. Second, without the artifact-free ground truth, a deep unsupervised MAR framework in the reverse process is constructed to contextually learn the known background data distribution for the unknown metal trace restoration in sinogram-domain. Third, to further maintain the sinogram-domain data consistency, two physics-based consistency constraint loss functions, including conjugate-ray and accumulation-ray consistency loss, are designed for the conjugate point constraint and the accumulation constraint. The proposed UPGRADE-Net is trained and evaluated on a publicly available dataset and a clinical dataset. Extensive experimental results validate that the proposed method outperforms the state-of-the-art competing methods for MAR.","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4201-4211"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145484671","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"FoundDiff: Foundational Diffusion Model for Generalizable Low-Dose CT Denoising","authors":"Zhihao Chen;Qi Gao;Zilong Li;Junping Zhang;Yi Zhang;Jun Zhao;Hongming Shan","doi":"10.1109/TMI.2026.3698474","DOIUrl":"10.1109/TMI.2026.3698474","url":null,"abstract":"Low-dose computed tomography (CT) denoising is crucial for reduced radiation exposure while ensuring diagnostically acceptable image quality. Despite significant advancements driven by deep learning (DL) in recent years, existing DL-based methods, typically trained on a specific dose level and anatomical region, struggle to handle diverse noise characteristics and anatomical heterogeneity during varied scanning conditions, limiting their generalizability and robustness in clinical scenarios. In this paper, we propose FoundDiff, a foundational diffusion model for unified and generalizable LDCT denoising across various dose levels and anatomical regions. FoundDiff employs a two-stage strategy: <bold>(i) </b> dose-anatomy perception and <bold>(ii)</b> adaptive denoising. First, we develop a dose- and anatomy-aware contrastive language-image pre-training model (DA-CLIP) to achieve robust dose and anatomy perception by leveraging specialized contrastive learning strategies to learn continuous representations that quantify ordinal dose variations and identify salient anatomical regions. Second, we design a dose- and anatomy-aware diffusion model (DA-Diff) to perform adaptive and generalizable denoising by synergistically integrating the learned dose and anatomy embeddings from DA-CLIP into diffusion process via a novel dose and anatomy conditional block (DACB) based on Mamba. Extensive experiments on a large simulated multi-dose CT dataset spanning three anatomical regions, together with cross-dataset evaluations on Mayo-2016, CQ500, and piglet datasets, demonstrate superior denoising performance and strong generalization to unseen dose levels and anatomical regions. The codes and models are available at <uri>https://github.com/hao1635/FoundDiff</uri>","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4366-4379"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148145477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"RVDeformer: Sparse Point Cloud-Guided Right Ventricle 3-D Reconstruction in Echocardiograms","authors":"Zhaohui Wang;Jun Shi;Minfan Zhao;Ziqi Zhu;Yida Li;Hong An","doi":"10.1109/TMI.2026.3696549","DOIUrl":"10.1109/TMI.2026.3696549","url":null,"abstract":"3D reconstruction of the Right Ventricle (RV) from echocardiograms is crucial for accurate clinical evaluation of cardiac function. However, existing methods are hindered by the complex RV anatomy and the incomplete spatial information inherent in 2D multi-view echocardiograms. Therefore, we propose RVDeformer, a sparse point cloud-guided framework for RV 3D reconstruction. RVDeformer reformulates the reconstruction task as a mesh deformation problem, learning to deform a predefined template mesh to match the target structure under the guidance of the sparse anatomical point cloud. Specifically, this framework employs the end-to-end neural network RVDeformNet to extract the features of the point cloud and template mesh for predicting the displacement of each mesh vertex. We design a point cloud-mesh fusion module that can effectively align and fuse features from the two modalities to enhance the representation ability of the model. We conduct extensive validation on a clinical dataset of 1,278 cases and demonstrate that RVDeformer outperforms existing state-of-the-art methods, achieving a Chamfer Distance (CD) of <inline-formula> <tex-math>$2.24pm 0.55$ </tex-math></inline-formula> mm, an F1-score of <inline-formula> <tex-math>$0.74pm 0.10$ </tex-math></inline-formula> at the 3 mm threshold, and a Volumetric Similarity (VS) of <inline-formula> <tex-math>$91.53pm 2.28$ </tex-math></inline-formula>%, with significant potential for clinical applications. The code is available at <uri>https://github.com/onezh95/RVDeformer</uri>","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4240-4253"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148019685","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Lesion Detectability and Masking Disparity Assessment in Breast Tomosynthesis Across Diverse Populations Using In-Silico Imaging Trials","authors":"Bruno Barufaldi;Rodrigo B. Vimieiro;Vincent Dong;Hanna Tomic;Quy Cao;Marcelo Andrade da Costa Vieira;Predrag R. Bakic;Peter R. Eby;Sophia Zackrisson;Anne Marie McCarthy;Andrew D. A. Maidment","doi":"10.1109/TMI.2026.3701599","DOIUrl":"10.1109/TMI.2026.3701599","url":null,"abstract":"Breast density impacts cancer detection by masking tumors within fibroglandular tissue and there are disparities in screening outcomes across racial groups. However, it remains unclear whether these differences reflect inherent tissue characteristics or systemic bias. To isolate the effect of breast density on lesion detectability across racial subgroups, we conducted a retrospective case-control study using raw tomosynthesis projections from 902 women (453 cases, 451 matched controls) across BI-RADS density categories and self-reported race. Identical <italic>in-silico</i> spiculated masses (8–15 mm) and microcalcification clusters (10–14 mm) were inserted into the projections using a calibrated lesion model. Images were reconstructed in the same manner to avoid proprietary processing in lesion detection. Lesion detectability was assessed with Channelized Hotelling Observers. Regression and causal mediation analyses examined the relationships between race, density, and detectability. As result, detectability decreased with increasing density; for masses, the area under the receiver operating characteristic curve (ROC AUC) reduced significantly from 0.93 to 0.85 (BIRADS A to D), whereas for microcalcifications AUC decreased from 0.85 to 0.78 across the same density range. Discrimination remained higher for masses than calcifications (AUC = 0.89 vs. AUC = 0.82). Stratified analyses showed slightly higher detectability in Non-Hispanic Black women compared with Non-Hispanic White and Asian American women, largely reflecting differences in density. Mediation analysis revealed that breast density accounted for 38–55% of the observed race-associated detectability differences. Mediation analyses have shown that density is the dominant factor in detectability. These findings support the development of calibrated detection models and personalized screening strategies that account for breast density.","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4433-4443"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148213921","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Boundary-Aware Spectral and Morphological Guidance Method for Feature-Driven Colorectal Cancer Segmentation","authors":"Pengquan Lei;Hengxiao Hu;Suyun Li;Yajun Liu;Xiaowen Xie;Jingfan Zhan;Kuanhong Wang;Yingying Guo;Zaiyi Liu;Bingjiang Qiu;Xin Chen","doi":"10.1109/TMI.2026.3705577","DOIUrl":"10.1109/TMI.2026.3705577","url":null,"abstract":"Precise segmentation of medical images plays a crucial role in modern clinical practice, providing important foundations for the quantitative analysis of medical images and clinical decision making. However, although deep learning techniques have achieved significant success in conventional medical image segmentation, they still exhibit obvious limitations when faced with complex structure segmentation tasks such as colorectal cancer: high-quality medical data acquisition is not only difficult, but also even when data is relatively sufficient, the diverse morphological features of lesions cannot be adequately represented due to their high variability, which severely limits the generalization capability of traditional data-driven segmentation methods. To address these challenges, we propose a feature-driven segmentation model that improves performance by deeply mining intrinsic data information rather than relying on parameter stacking. Specifically, we introduce a spectrum–prior–boundary triple modeling paradigm, where frequency domain reconstruction and modulation is employed to establish mappings between different frequency bands and heterogeneous signals to identify ambiguous signals, a level set–based segmentation algorithm is used to construct abdominal anatomical distance fields to incorporate morphological priors, and an auxiliary edge branch is designed by integrating deep semantic and shallow detail features to strengthen boundary awareness. Extensive experiments on colorectal cancer segmentation demonstrate that the proposed method achieves significant improvements in both segmentation accuracy and boundary perception over other state-of-the-art methods. Furthermore, evaluations on lung cancer and breast cancer segmentation tasks validate its strong generalization capability across diverse lesion types.","PeriodicalId":94033,"journal":{"name":"IEEE transactions on medical imaging","volume":"45 8","pages":"4559-4578"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148298909","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}