Neural implicit k-space representations (NIK) have shown promising results for dynamic magnetic resonance imaging (MRI) at high temporal resolutions. Yet, reducing acquisition time, and thereby available training data, results in severe performance drops due to overfitting. To address this, we introduce a novel self-supervised k-space loss function L PISCO , applicable for regularization of NIK-based reconstructions. The proposed loss function is based on the concept of parallel imaging-inspired self-consistency (PISCO), enforcing a consistent global k-space neighborhood relationship without requiring additional data. Quantitative and qualitative evaluations on static and dynamic MR reconstructions show that integrating PISCO significantly improves NIK representations, making it a competitive dynamic reconstruction method without constraining the temporal resolution. Particularly at high acceleration factors (R ≥ 50), NIK with PISCO can avoid temporal oversmoothing of state-of-the-art methods and achieves superior spatio-temporal reconstruction quality. Furthermore, an extensive analysis of the loss assumptions and stability shows PISCO’s potential as versatile self-supervised k-space loss function for further applications and architectures. Code is available at: https://github.com/compai-lab/2025-pisco-spieker

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.

In MR Fingerprinting (MRF), balanced Steady-State Free Precession (bSSFP) has advantages over unbalanced SSFP because it retains the spin history achieving a higher signal-to-noise ratio (SNR) and scan efficiency. However, bSSFP-MRF is not frequently used because it is sensitive to off-resonance, producing artifacts and blurring, and affecting the parametric map quality. Here we propose a novel Spatial Off-resonance Correction (SOC) approach for reducing these artifacts in bSSFP-MRF with spiral trajectories. SOC-MRF uses each pixel's Point Spread Function to create system matrices that encode both off-resonance and gridding effects. We iteratively compute the inverse of these matrices to reduce the artifacts. We evaluated the proposed method using brain simulations and actual MRF acquisitions of a standardized T1/T2 phantom and five healthy subjects. The results show that the off-resonance distortions in T1/T2 maps were considerably reduced using SOC-MRF. For T2, the Normalized Root Mean Square Error (NRMSE) was reduced from 17.3 to 8.3% (simulations) and from 35.1 to 14.9% (phantom). For T1, the NRMS was reduced from 14.7 to 7.7% (simulations) and from 17.7 to 6.7% (phantom). For in-vivo, the mean and standard deviation in different ROI in white and gray matter were significantly improved. For example, SOC-MRF estimated an average T2 for white matter of 77ms (the ground truth was 74ms) versus 50 ms of MRF. For the same example the standard deviation was reduced from 18 ms to 6ms. The corrections achieved with the proposed SOC-MRF may expand the potential applications of bSSFP-MRF, taking advantage of its better SNR property.

Physiological motion, such as cardiac and respiratory motion, during Magnetic Resonance (MR) image acquisition can cause image artifacts. Motion correction techniques have been proposed to compensate for these types of motion during thoracic scans, relying on accurate motion estimation from undersampled motion-resolved reconstruction. A particular interest and challenge lie in the derivation of reliable non-rigid motion fields from the undersampled motion-resolved data. Motion estimation is usually formulated in image space via diffusion, parametric-spline, or optical flow methods. However, image-based registration can be impaired by remaining aliasing artifacts due to the undersampled motion-resolved reconstruction. In this work, we describe a formalism to perform non-rigid registration directly in the sampled Fourier space, i.e. k-space. We propose a deep-learning based approach to perform fast and accurate non-rigid registration from the undersampled k-space data. The basic working principle originates from the Local All-Pass (LAP) technique, a recently introduced optical flow-based registration. The proposed LAPNet is compared against traditional and deep learning image-based registrations and tested on fully-sampled and highly-accelerated (with two undersampling strategies) 3D respiratory motion-resolved MR images in a cohort of 40 patients with suspected liver or lung metastases and 25 healthy subjects. The proposed LAPNet provided consistent and superior performance to image-based approaches throughout different sampling trajectories and acceleration factors.

Non-rigid motion-corrected reconstruction has been proposed to account for the complex motion of the heart in free-breathing 3D coronary magnetic resonance angiography (CMRA). This reconstruction framework requires efficient and accurate estimation of non-rigid motion fields from undersampled images at different respiratory positions (or bins). However, state-of-the-art registration methods can be time-consuming. This article presents a novel unsupervised deep learning-based strategy for fast estimation of inter-bin 3D non-rigid respiratory motion fields for motion-corrected free-breathing CMRA. The proposed 3D respiratory motion estimation network (RespME-net) is trained as a deep encoder-decoder network, taking pairs of 3D image patches extracted from CMRA volumes as input and outputting the motion field between image patches. Using image warping by the estimated motion field, a loss function that imposes image similarity and motion smoothness is adopted to enable training without ground truth motion field. RespME-net is trained patch-wise to circumvent the challenges of training a 3D network volume-wise which requires large amounts of GPU memory and 3D datasets. We perform 5-fold cross-validation with 45 CMRA datasets and demonstrate that RespME-net can predict 3D non-rigid motion fields with subpixel accuracy (0.44 ± 0.38 mm) within ~10 seconds, being ~20 times faster than a GPU-implemented state-of-the-art non-rigid registration method. Moreover, we perform non-rigid motion-compensated CMRA reconstruction for 9 additional patients. The proposed RespME-net has achieved similar motion-corrected CMRA image quality to the conventional registration method regarding coronary artery length and sharpness.

Manifold alignment (MA) is a technique to map many high-dimensional datasets to one shared low-dimensional space. Here we develop a pipeline for using MA to reconstruct high-resolution medical images. We present two key contributions. First, we develop a novel MA scheme in which each high-dimensional dataset can be differently weighted preventing noisier or less informative data from corrupting the aligned embedding. We find that this generalisation improves performance in our experiments in both supervised and unsupervised MA problems. Second, we use the wave kernel signature as a graph descriptor for the unsupervised MA case finding that it significantly outperforms the current state-of-the-art methods and provides higher quality reconstructed magnetic resonance volumes than existing methods.

Segmenting anatomical structures in medical images has been successfully addressed with deep learning methods for a range of applications. However, this success is heavily dependent on the quality of the image that is being segmented. A commonly neglected point in the medical image analysis community is the vast amount of clinical images that have severe image artefacts due to organ motion, movement of the patient and/or image acquisition related issues. In this paper, we discuss the implications of image motion artefacts on cardiac MR segmentation and compare a variety of approaches for jointly correcting for artefacts and segmenting the cardiac cavity. The method is based on our recently developed joint artefact detection and reconstruction method, which reconstructs high quality MR images from k-space using a joint loss function and essentially converts the artefact correction task to an under-sampled image reconstruction task by enforcing a data consistency term. In this paper, we propose to use a segmentation network coupled with this in an end-to-end framework. Our training optimises three different tasks: 1) image artefact detection, 2) artefact correction and 3) image segmentation. We train the reconstruction network to automatically correct for motion-related artefacts using synthetically corrupted cardiac MR k-space data and uncorrected reconstructed images. Using a test set of 500 2D+time cine MR acquisitions from the UK Biobank data set, we achieve demonstrably good image quality and high segmentation accuracy in the presence of synthetic motion artefacts. We showcase better performance compared to various image correction architectures.

Good quality of medical images is a prerequisite for the success of subsequent image analysis pipelines. Quality assessment of medical images is therefore an essential activity and for large population studies such as the UK Biobank (UKBB), manual identification of artefacts such as those caused by unanticipated motion is tedious and time-consuming. Therefore, there is an urgent need for automatic image quality assessment techniques. In this paper, we propose a method to automatically detect the presence of motion-related artefacts in cardiac magnetic resonance (CMR) cine images. We compare two deep learning architectures to classify poor quality CMR images: 1) 3D spatio-temporal Convolutional Neural Networks (3D-CNN), 2) Long-term Recurrent Convolutional Network (LRCN). Though in real clinical setup motion artefacts are common, high-quality imaging of UKBB, which comprises cross-sectional population data of volunteers who do not necessarily have health problems creates a highly imbalanced classification problem. Due to the high number of good quality images compared to the relatively low number of images with motion artefacts, we propose a novel data augmentation scheme based on synthetic artefact creation in k-space. We also investigate a learning approach using a predetermined curriculum based on synthetic artefact severity. We evaluate our pipeline on a subset of the UK Biobank data set consisting of 3510 CMR images. The LRCN architecture outperformed the 3D-CNN architecture and was able to detect 2D+time short axis images with motion artefacts in less than 1ms with high recall. We compare our approach to a range of state-of-the-art quality assessment methods. The novel data augmentation and curriculum learning approaches both improved classification performance achieving overall area under the ROC curve of 0.89.

In this paper, we propose a generalized joint sparsity regularization prior and reconstruction framework for the synergistic reconstruction of positron emission tomography (PET) and under sampled sensitivity encoded magnetic resonance imaging data with the aim of improving image quality beyond that obtained through conventional independent reconstructions. The proposed prior improves upon the joint total variation (TV) using a non-convex potential function that assigns a relatively lower penalty for the PET and MR gradients, whose magnitudes are jointly large, thus permitting the preservation and formation of common boundaries irrespective of their relative orientation. The alternating direction method of multipliers (ADMM) optimization framework was exploited for the joint PET-MR image reconstruction. In this framework, the joint maximum a posteriori objective function was effectively optimized by alternating between well-established regularized PET and MR image reconstructions. Moreover, the dependency of the joint prior on the PET and MR signal intensities was addressed by a novel alternating scaling of the distribution of the gradient vectors. The proposed prior was compared with the separate TV and joint TV regularization methods using extensive simulation and real clinical data. In addition, the proposed joint prior was compared with the recently proposed linear parallel level sets (PLSs) method using a benchmark simulation data set. Our simulation and clinical data results demonstrated the improved quality of the synergistically reconstructed PET-MR images compared with the unregularized and conventional separately regularized methods. It was also found that the proposed prior can outperform both the joint TV and linear PLS regularization methods in assisting edge preservation and recovery of details, which are otherwise impaired by noise and aliasing artifacts. In conclusion, the proposed joint sparsity regularization within the presented a ADMM reconstruction framework is a promising technique, nonetheless our clinical results showed that the clinical applicability of joint reconstruction might be limited in current PET-MR scanners, mainly due to the lower resolution of PET images.

We present a novel retrospective self-gating method based on manifold alignment (MA), which enables reconstruction of free breathing, high spatial, and temporal resolution abdominal magnetic resonance imaging sequences. Based on a radial golden-angle acquisition trajectory, our method enables a multidimensional self-gating signal to be extracted from the k -space data for more accurate motion representation. The k -space radial profiles are evenly divided into a number of overlapping groups based on their radial angles. MA is then used to simultaneously learn and align the low dimensional manifolds of all groups, and embed them into a common manifold. In the manifold, k -space profiles that represent similar respiratory positions are close to each other. Image reconstruction is performed by combining radial profiles with evenly distributed angles that are close in the manifold. Our method was evaluated on both 2-D and 3-D synthetic and in vivo data sets. On the synthetic data sets, our method achieved high correlation with the ground truth in terms of image intensity and virtual navigator values. Using the in vivo data, compared with a state-of-the-art approach based on the center of k -space gating, our method was able to make use of much richer profile data for self-gating, resulting in statistically significantly better quantitative measurements in terms of organ sharpness and image gradient entropy.

We present a novel framework for retrospective dynamic 3D volume reconstruction from a multi-slice MRI acquisition using manifold alignment. K-space data are continuously acquired under free breathing using a radial golden-angle trajectory in a slice-by-slice manner. Non-overlapping consecutive profiles that were acquired within a short time window are grouped together. All grouped profiles from all slices are then simultaneously embedded using manifold alignment into a common manifold space (MS), in which profiles that were acquired at similar respiratory states are close together. Subsequently, a 3D volume can be reconstructed at each of the grouped profile MS positions by combining profiles that are close in the MS. This enables the original multi-slice dataset to be used to reconstruct a dynamic 3D sequence based on the respiratory state correspondences established in the MS. Our method was evaluated on both synthetic and in vivo datasets. For the synthetic datasets, the reconstructed dynamic sequence achieved a normalised cross correlation of 0.98 and peak signal to noise ratio of 26.64 dB compared with the ground truth. For the in vivo datasets, based on sharpness measurements and visual comparison, our method performed better than reconstruction using an adapted central k-space gating method.

Motion occurring during magnetic resonance imaging acquisition is a major factor of image quality degradation. Self-navigation can help reduce artefacts by estimating motion from the acquired data to enable motion correction. Popular self-navigation techniques rely on the availability of a fully-sampled motion-free reference to register the motion corrupted data with. In the proposed technique, rigid motion parameters are derived using the inherent correlation between radial segments in k-space. The registration is performed exclusively in k-space using the Phase Correlation Method, a popular registration technique in computer vision. Robust and accurate registration has been carried out from radial segments composed of as few as 32 profiles. Successful self-navigation has been performed on 2-D dynamic brain scans corrupted with continuous motion for six volunteers. Retrospective motion correction using the derived self-navigation parameters resulted in significant improvement of image quality compared to the conventional sliding window. This work also demonstrates the benefits of using a bit-reversed ordering scheme to limit undesirable effects specific to retrospective motion correction on radial trajectories. This method provides a fast and efficient mean of measuring rigid motion directly in k-space from dynamic radial data under continuous motion.

BACKGROUND: Calcium (Ca2+) propagates within tissues serving as an important information carrier. In particular, cilia beat frequency in oviduct cells is partially regulated by Ca2+ changes. Thus, measuring the calcium density and characterizing the traveling wave plays a key role in understanding biological phenomena. However, current methods to measure propagation velocities and other wave characteristics involve several manual or time-consuming procedures. This limits the amount of information that can be extracted, and the statistical quality of the analysis. RESULTS: Our work provides a framework based on image processing procedures that enables a fast, automatic and robust characterization of data from two-filter fluorescence Ca2+ experiments. We calculate the mean velocity of the wave-front, and use theoretical models to extract meaningful parameters like wave amplitude, decay rate and time of excitation. CONCLUSIONS: Measurements done by different operators showed a high degree of reproducibility. This framework is also extended to a single filter fluorescence experiments, allowing higher sampling rates, and thus an increased accuracy in velocity measurements.

Magnetic resonance imaging (MRI) has been commonly used for guiding and planning image guided interventions since it provides excellent soft tissue visualization of anatomy and allows motion modeling to predict the position of target tissues during the procedure. However, MRI-based motion modeling remains challenging due to the difficulty of acquiring multiple motion-free 3-D respiratory phases with adequate contrast and spatial resolution. Here, we propose a novel retrospective respiratory gating scheme from a 3-D undersampled high-resolution MRI acquisition combined with fast and robust image registrations to model the nonrigid deformation of the liver. The acquisition takes advantage of the recently introduced golden-radial phase encoding (G-RPE) trajectory. G-RPE is self-gated, i.e., the respiratory signal can be derived from the acquired data itself, and allows retrospective reconstructions of multiple respiratory phases at any arbitrary respiratory position. Nonrigid motion modeling is applied to predict the liver deformation of an average breathing cycle. The proposed approach was validated on 10 healthy volunteers. Motion model accuracy was assessed using similarity-, surface-, and landmark-based validation methods, demonstrating precise model predictions with an overall target registration error of TRE = 1.70 ± 0.94 mm which is within the range of the acquired resolution.

Long acquisition times are still a limitation for many applications of magnetic resonance imaging (MRI), specially in 3-D and dynamic imaging. Several undersampling reconstruction techniques have been proposed to overcome this problem. These techniques are based on acquiring less samples than specified by the Nyquist criterion and estimating the nonacquired data by using some sort of prior information. Most of these reconstruction methods use prior information based on estimations of the pixel intensities of the images and therefore they are prone to introduce spatial or temporal blurring. Instead of using the pixel intensities, we propose to use information that allows us to sort the pixels of an image from darkest to brightest. The set of order relations which sort the pixels of an image has been called intensity order. The intensity order of an image can be estimated from low-resolution images, adjacent slices in volumetric acquisitions, temporal correlation in dynamic sequences or from prior reconstructions. Our technique for reconstruction using intensity order (TRIO) consists of looking for an image that satisfies the intensity order and minimizes the discrepancy between the acquired and reconstructed data. Results show that TRIO can effectively reconstruct 2-D-cine cardiac MR images (under-sampling factor of 4), estimating correctly the temporal evolution of the objects. Furthermore, TRIO is used as a second stage reconstruction after reconstructing with other techniques, keyhole, sliding window and k-t BLAST, to estimate the order information. In all cases the images are improved by TRIO.

Image segmentation has attracted the attention of researcher for many decades. Different approaches have been developed in order to find the solution in many different segmentation situations. In this paper we propose a novel edge detection approach aimed to generate useful information to achieve segmentation. The proposed method is based on analysis of the information provided by the time matrix generated from a pulse coupled neural network, PCNN. This information represents gray level differences among the pixel images. Two different schemes for edge detection are presented. The first scheme is developed to generate edges from coarse images and the second one to deal with more detailed edges. Similarity of this method with a previous developed method based on fuzzy edge level detection is also covered in the paper. Final results show that the proposed method may be used as a new alternative to define image edges of different levels for further analysis.