The low degree of labeling and limited photon count of fluorescent emitters in single molecule localization microscopy results in poor quality images of macro-molecular complexes. Particle fusion provides a single reconstruction with high signal-to-noise ratio by combining many single molecule localization microscopy images of the same structure. The underlying assumption of homogeneity is not always valid, heterogeneity can arise due to geometrical shape variations or distinct conformational states. We introduce a Point Cloud Variational Auto-Encoder that works directly on 2D and 3D localization data, to detect multiple modes of variation in such datasets. The computing time is on the order of a few minutes, enabled by the linear scaling with dataset size, and fast network training in just four epochs. The use of lists of localization data instead of pixelated images leads to just minor differences in computational burden between 2D and 3D cases. With the proposed method, we detected radius variation in 2D Nuclear Pore Complex data, height variations in 3D DNA origami tetrahedron data, and both radius and height variations in 3D Nuclear Pore Complex data. In all cases, the detected variations were on the few nanometer scale.

Pretrained segmentation models for cardiac magnetic resonance imaging (MRI) struggle to generalize across different imaging sequences due to significant variations in image contrast. These variations arise from changes in imaging protocols, yet the same fundamental spin properties, including proton density, T₁, and T₂ values, govern all acquired images. With this core principle, we introduce Reverse Imaging, a novel physics-driven method for cardiac MRI data augmentation and domain adaptation to fundamentally solve the generalization problem. Our method reversely infers the underlying spin properties from observed cardiac MRI images, by solving ill-posed nonlinear inverse problems regularized by the prior distribution of spin properties. We acquire this “spin prior” by learning a generative diffusion model from the multiparametric SAturation-recovery single-SHot acquisition sequence (mSASHA) dataset, which offers joint cardiac T₁ and T₂ maps. Our method enables approximate but meaningful spin-property estimates from MR images, which provide an interpretable “latent variable” that lead to highly flexible image synthesis of arbitrary novel sequences. We show that Reverse Imaging enables highly accurate segmentation across vastly different image contrasts and imaging protocols, realizing wide-spectrum generalization of cardiac MRI segmentation.

Background: As a quantitative magnetic resonance imaging (MRI) technique, myocardial T1 mapping plays a crucial role in the diagnosis and treatment of cardiovascular diseases. In practice, involuntary cardiac and respiratory motion often results in reduced accuracy and precision in T1 estimation. Therefore, image registration remains crucial for accurate and precise myocardial T1 mapping. Compared with pairwise registration that warps each baseline image to a predefined template, groupwise registration aligns all images from one sequence simultaneously without the need for a template. However, a persistent challenge is the difficulty of extracting the structural representation of T1 mapping data that contains vastly varying contrast, which severely undermines the performance of image registration. Purpose: The purpose of this study is to incorporate the learning capabilities of the diffusion model to tackle the main challenge encountered in the registration of myocardial T1 mapping. Our goal is to align all images within an image series simultaneously in a groupwise manner. Methods: In this article, we propose a novel template-free groupwise registration framework that can align one T1-weighted image series through a single forward propagation. Notably, we introduce the diffusion process to effectively boost the structural information extraction under the drastic contrast changes for reliable image registration. Furthermore, we design a Hybrid Attention Feature Fusion (HAFF) module to promote the multi-scale feature fusion from diffusion to registration. To evaluate the registration performance of the proposed model, experiments are conducted on a publicly available myocardial T1 mapping dataset comprising 210 consecutive patients, using an independent test set for comparison experiments and ablation studies. Results: Experimental results demonstrated the great superiority of our proposed method in the registration of myocardial T1 mapping. Quantitatively, the proposed method resulted in a Dice score of 0.839, groupwise Dice score of 0.601, Hausdorff distance of 10.389 mm, and T1 mapping error of 11.372 ms, surpassing the current state-of-the-art approaches. Conclusions: Our proposed framework realizes robust groupwise registration for myocardial T1 mapping by leveraging the state-of-the-art diffusion model, demonstrating its strong feature extraction capacity for image registration, beyond image generation.

Deformable image registration remains a central challenge in medical image analysis, particularly under multi-modal scenarios where intensity distributions vary significantly across scans. While deep learning methods provide efficient feed-forward predictions, they often fail to generalize robustly under distribution shifts at test time. A straightforward remedy is full network fine-tuning, yet for modern architectures such as Transformers or deep U-Nets, this adaptation is prohibitively expensive in both memory and runtime when operating in 3D. Meanwhile, the naive fine-tuning struggles more with potential degradation in performance in the existence of drastic domain shifts. In this work, we propose a registration framework that integrates a frozen pretrained mono-modal registration model with a lightweight adaptation pipeline for multi-modal image registration. Specifically, we employ style transfer based on contrast-agnostic representation generation and refinement modules to bridge modality and domain gaps with instance optimization at test time. This design is orthogonal to the choice of backbone mono-modal model, thus avoids the computational burden of full fine-tuning while retaining the flexibility to adapt to unseen domains. We evaluate our approach on the Learn2Reg 2025 LUMIR validation set and observe consistent improvements over the pretrained state-of-the-art mono-modal backbone. In particular, the method ranks second on the multi-modal subset, third on the out-of-domain subset, and achieves fourth place overall in Dice score. These results demonstrate that combining frozen mono-modal models with modality adaptation and lightweight instance optimization offers an effective and practical pathway toward robust multi-modal registration.

Multiparametric mapping MRI has become a viable tool for myocardial tissue characterization. However, misalignment between multiparametric maps makes pixel-wise analysis challenging. To address this challenge, we developed a generalizable physics-informed deep-learning model using test-time adaptation to enable group image registration across contrast weighted images acquired from multiple physical models (e.g., a T1 mapping model and T2 mapping model). The physics-informed adaptation utilized the synthetic images from specific physics model as registration reference, allows for transductive learning for various tissue contrast. We validated the model in healthy volunteers with various MRI sequences, demonstrating its improvement for multi-modal registration with a wide range of image contrast variability.

Objectives: To evaluate a deep learning (DL) model for reducing the agent dose of contrast-enhanced T1-weighted MRI (T1ce) of the cerebellopontine angle (CPA) cistern. Materials and methods: In this multicenter retrospective study, T1 and T1ce of vestibular schwannoma (VS) patients were used to simulate low-dose T1ce with varying reductions of contrast agent dose. DL models were trained to restore standard-dose T1ce from the low-dose simulation. The image quality and segmentation performance of the DL-restored T1ce were evaluated. A head and neck radiologist was asked to rate DL-restored images in multiple aspects, including image quality and diagnostic characterization. Results: 203 MRI studies from 72 VS patients (mean age, 58.51 ± 14.73, 39 men) were evaluated. As the input dose increased, the structural similarity index measure of the restored T1ce increased from 0.639 ± 0.113 to 0.993 ± 0.009, and the peak signal-to-noise ratio increased from 21.6 ± 3.73 dB to 41.4 ± 4.84 dB. At a 10% input dose, using DL-restored T1ce for segmentation improved the Dice from 0.673 to 0.734, the 95% Hausdorff distance from 2.38 mm to 2.07 mm, and the average surface distance from 1.00 mm to 0.59 mm. Both DL-restored T1ce from 10% and 30% input doses showed excellent image quality (3, interquartile range (IQR) [Q3-Q1] = 3–3 and 3, IQR [Q3-Q1] = 4–3), with the latter being considered more informative (4, IQR [Q3-Q1] = 4–3). Conclusion: The DL model improved the image quality of low-dose MRI of the CPA cistern, which makes lesion detection and diagnostic characterization possible with 10–30% of the standard dose. Key Points: Question Deep learning models that aid in the reduction of contrast agent dose are not extensively evaluated for MRI of the cerebellopontine angle cistern. Findings Deep learning models restored the low-dose MRI of the cerebellopontine angle cistern, yielding images sufficient for vestibular schwannoma diagnosis and management. Clinical relevance Deep learning models make it possible to reduce the use of gadolinium-based contrast agents for contrast-enhanced MRI of the cerebellopontine angle cistern.

Cardiac T1 mapping by magnetic resonance imaging (MRI) is an important clinical tool for the diagnosis and treatment of cardiovascular diseases. In practice, involuntary cardiac and respiratory motion often results in reduced accuracy and precision in T1 estimation. Motion correction is an essential preprocessing step, however, with intensive contrast changes among baseline images, both optimization-based and deep-learning (DL)-based registration methods still struggle to estimate structural similarity between images, especially when image contrast is poor and displacement is large. In this work, we propose a novel registration metric that is highly insensitive to large contrast changes, based on modified modality independent neighborhood descriptor (mo-MIND). To accommodate severe motions, we further propose pre-deformation as an augmentation strategy at the training stage. We combine the proposed mo-MIND-based metric and the augmentation strategy in a U-Net architecture to tackle the challenges of motion correction for cardiac T1 mapping. Experimental results and ablation studies demonstrated that our method achieved improved registration performance compared to several established baselines, leading to significantly reduced T1 mapping error and improved landmark stability.

Cardiovascular health is vital to human well-being, and cardiac magnetic resonance (CMR) imaging is considered the clinical reference standard for diagnosing cardiovascular disease. However, its adoption is hindered by long scan times, complex contrasts, and inconsistent quality. While deep learning methods perform well on specific CMR imaging sequences, they often fail to generalize across modalities and sampling schemes. The lack of benchmarks for high-quality, fast CMR image reconstruction further limits technology comparison and adoption. The CMRxRecon2024 challenge, attracting over 200 teams from 18 countries, addressed these issues with two tasks: generalization to unseen modalities and robustness to diverse undersampling patterns. We introduced the largest public multi-modality CMR raw dataset, an open benchmarking platform, and shared code. Analysis of the best-performing solutions revealed that prompt-based adaptation and enhanced physics-driven consistency enabled strong cross-scenario performance. These findings establish principles for generalizable reconstruction models and advance clinically translatable AI in cardiovascular imaging.

Dynamic brain data are becoming increasingly accessible, providing a gateway to understanding the inner workings of the brain in living participants. However, the size and complexity of the data pose a challenge in extracting meaningful information across various data sources. Here we introduce a generalizable unsupervised deep manifold learning for exploration of neurocognitive and behavioral patterns. Unlike existing methods that extract patterns directly from the input data, the proposed brain-dynamic convolutional-network-based embedding (BCNE) captures brain-state trajectories by analyzing temporospatial correlations within the data and applying manifold learning. The results demonstrate that BCNE effectively delineates scene transitions, underscores the involvement of different brain regions in memory and narrative processing, distinguishes dynamic learning processes and identifies differences between active and passive behaviors. BCNE provides an effective tool for exploring general neuroscience inquiries or individual-specific patterns.

Purpose of Review: This review explores the advancements in deep learning (DL)-based cardiac magnetic resonance (CMR) reconstruction, focusing on its role in accelerating imaging, denoising, super-resolution, motion artifact correction, and quantitative mapping. It highlights the transition from parallel imaging and compressed sensing to artificial intelligence (AI)-driven approaches that enhance image quality and diagnostic accuracy. Recent Findings: Supervised and self-supervised DL models can significantly reduce scan times, enabling high-fidelity reconstructions from undersampled data. Generative adversarial network (GAN)-based super-resolution techniques enhance spatial resolution, while denoising networks improve signal-to-noise ratio. Motion correction strategies, including spatiotemporal learning, have enhanced free-breathing acquisitions. Physics-guided models incorporate MRI signal constraints for improved T1/T2 mapping and myocardial tissue characterization. Summary: DL-driven CMR reconstruction optimizes imaging speed, quality, and artifact suppression. Despite challenges in dataset standardization and clinical validation, AI is advancing real-time, high-fidelity CMR, facilitating broader clinical adoption.