Parkinson's disease (PD) is a neurodegenerative disorder characterized by a range of motor and non-motor symptoms. Despite extensive research, the neural structure of time distortion, remains unclear. This study aimed to determine the neurobiological origins of time distortion by analyzing dynamic features in PD patients compared to control participants.
Approach.
We used PD and control electroencephalography (EEG) signals to investigate brain function during time distortion. The EEG signal was recorded during an interval-timing task. Following artifact reduction and EEG signal segmentation, dynamic features were extracted from each frequency band across all the channels. Channels showing significant discrepancies between the two groups were selected by statistical analysis. The features are sent to an Artificial Neural Network (ANN) classifier to evaluate their discriminative potential.
Main results.
The results indicated lower values of Lyapunov Exponent and Approximate Entropy along with higher value of Fractal dimension in PD which presented higher level of irregularity and randomness, particularly in the CPz, P5, P6, and C5 channels. The ANN classifier achieved 90% accuracy, 89% sensitivity, 87% F1 score, and 95% specificity in a 10-fold cross-validation. 
Significance.
Significantly different channels were concentrated in the central and parietal areas of the brain and were linked to decision-making, maintenance, and retrieval of stored information, and working memory. Moreover, based on dynamic EEG analysis, it seems that disrupted connections between the basal ganglia and posterior parietal cortex in PD appear to compromise frontoparietal network dysfunction, which is associated with impaired temporal processing in patients with PD.&#xD.

Despite major advances in dual-energy CT, obtaining accurate attenuation values for quantitative applications remains a technical challenge. To address this topic, we introduce a novel projection data domain material decomposition method that is an extension of an approach we recently proposed for beam hardening correction in single energy CT.&#xD;Approach. The proposed method employs object-specific scatter correction and an analytical energy response model. We compare its performance to image-based material decomposition on accuracy of attenuation values using the ACR-CT accreditation phantom, scanned with consecutive low and high energy axial scans in centered and off-centered positions. Accuracy is assessed across the five inserts, and the images are analyzed for beam hardening artifacts and noise. Additionally, we assess the usefulness of object-specific scatter correction, and we assess performance over conventional data domain material decomposition and for anthropomorphic abdomen phantom imaging. &#xD;Main results. In the ACR phantom, the proposed method yielded a significant improvement in accuracy of the attenuation values, particularly at low energies (< 70keV), and an important reduction in beam hardening artifacts. While similarly high accuracy was achieved for water, quantitative error within the non-water inserts was lower and more uniform across the 30-140keV range, especially in the more challenging off-centered positioning of the phantom. Noise showed expected parabolic behavior, but with minimum at lower keV, which may be clinically advantageous. Object-specific scatter correction was shown to prevent major artifacts. Advantages over conventional data-domain decomposition clearly appeared when only a standard phantom is available to calibrate the latter. Lastly, the proposed method was shown to perform well, without any changes, in the more complex scenario of abdominal phantom imaging. &#xD;Significance. This work demonstrates that data-based material decomposition using an analytical energy response model with object-specific scatter correction offers a promising pathway to improve accuracy of CT attenuation values.

Objective.Wearable devices with embedded photoplethysmography (PPG) enable continuous non-invasive monitoring of cardiac activity, offering a promising strategy to reduce the global burden of cardiovascular diseases. However, monitoring during daily life introduces motion artifacts that can compromise the signals. Traditional signal decomposition techniques often fail with severe artifacts. Deep learning denoisers are more effective but have poorer interpretability, which is critical for clinical acceptance. This study proposes a framework that combines the advantages of both signal decomposition and deep learning approaches.Approach.We leverage algorithm unfolding to integrate prior knowledge about the PPG structure into a deep neural network, improving its interpretability. A learned convolutional sparse coding model encodes the signal into a sparse representation using a learned dictionary of kernels that capture recurrent morphological patterns. The network is trained for denoising using the PulseDB dataset and a synthetic motion artifact model from the literature. Performance is benchmarked with PPG during daily activities using the PPG-DaLiA dataset and compared with two reference deep learning methods.Main results.On the synthetic dataset, the proposed method, on average, improved the signal-to-noise ratio (SNR) from -7.06 dB to 11.23 dB and reduced the heart rate mean absolute error (MAE) by 55%. On the PPG-DaLiA dataset, the MAE decreased by 23%. The proposed method obtained higher SNR and comparable MAE to the reference methods.Significance.Our method effectively enhances the quality of PPG signals from wearable devices and enables the extraction of meaningful waveform features, which may inspire innovative tools for monitoring cardiovascular diseases.

Obtaining multiple CT scans from the same patient is required in many clinical scenarios, such as lung nodule screening and image-guided radiation therapy. Repeated scans would expose patients to higher radiation dose and increase the risk of cancer. In this study, we aim to achieve ultra-low-dose imaging for subsequent scans by collecting extremely undersampled sinogram via regional few-view scanning, and preserve image quality utilizing the preceding fullsampled scan as prior. To fully exploit prior information, we propose a two-stage framework consisting of diffusion model-based sinogram restoration and deep learning-based unrolled iterative reconstruction. Specifically, the undersampled sinogram is first restored by a conditional diffusion model with sinogram-domain prior guidance. Then, we formulate the undersampled data reconstruction problem as an optimization problem combining fidelity terms for both undersampled and restored data, along with a regularization term based on image-domain prior. Next, we propose Prior-aided Alternate Iterative NeTwork (PAINT) to solve the optimization problem. PAINT alternately updates the undersampled or restored data fidelity term, and unrolls the iterations to integrate neural network-based prior regularization. In the case of 112 mm field of view in simulated data experiments, our proposed framework achieved superior performance in terms of CT value accuracy and image details preservation. Clinical data experiments also demonstrated that our proposed framework outperformed the comparison methods in artifact reduction and structure recovery.

Prompt guiding multi-scale adaptive sparse representation-driven network for low-dose CT MAR.

This study was aimed at identifying MRI findings related to total hip arthroplasty (THAs) infection using coronal STIR with metal artifact reduction sequences (MARS) at 1.5T. This retrospective multicenter study included all patients with THAs who underwent 1.5T MRI with MARS from December 2015 to April 2020. Two groups are as follows: an infected group and a non-infected group (including asymptomatic THAs and symptomatic non-infected THAs). MARS were either multi-acquisition with variable-resonance image combination (MAVRIC) or slice encoding for metal artifact correction (SEMAC). Imaging features were evaluated to assess their association with THA infection (including both symptomatic and asymptomatic patients). Sensitivity, specificity, and accuracy of these imaging findings were assessed, and inter-reader agreement (kappa, K) was determined. Sixteen patients with THAs had periprosthetic infection, compared with 46 THAs in the non-infected group. Bone edema extending to adjacent soft tissues, defined as a combination of femoral bone marrow edema, hyperintense cortical signal, periostitis, and overall soft tissue edema, had the greatest diagnostic performance for infection: 15/16(94%) infected THAs and 0/46(0%) non-infected THAs (accuracy = 0.98, sensitivity = 0.94, specificity = 1, p < 0.001 Chi-Square test). This combination, predominant in the infected group (p < 0.001 for all), also demonstrated separately high accuracy (acc = 0.94-1), sensitivity (se = 0.94-1), and specificity (0.94-1). Fistula and fluid collection were highly specific (spe = 1) and accurate (acc = 0.81-0.82), although less sensitive (se = 0.25-0.31, p < 0.001). The combination of femoral bone marrow edema, hyperintense cortical signal, periostitis, and soft tissue edema is accurate in the diagnosis of periprosthetic hip joint infection using coronal STIR with MARs at 1.5T.

Objective.Normalized metal artifact reduction (NMAR) is a robust and widely used method for reducing metal artifacts in computed tomography (CT). However, conventional NMAR requires at least two forward projections, one for metal trace detection and the other for prior sinogram generation, resulting in redundant computation and limited efficiency. This study aims to reformulate NMAR into a single forward projection-based framework that maintains artifact reduction performance while improving computational efficiency and structural simplicity.Approach.We show that the two separate forward projections in NMAR can be unified into a single operation by leveraging deep learning (DL) priors, thereby eliminating the explicit forward projection for metal trace. The metal trace is inferred directly from localized discrepancies between the original sinogram and the forward projection of the DL prior image, allowing both interpolation and trace identification within a unified forward projection. Simulations and cadaver experiments were performed to compare the proposed method with NMAR, DL reconstruction, and conventional DL-NMAR.Main results.The proposed method reduced metal artifacts with image quality comparable to conventional DL-NMAR while improving computational efficiency. By reducing the number of forward projections from two to one, the proposed method achieved the lowest number of projection operations among all compared methods, highlighting its computational advantage.Significance.This study demonstrates that DL priors can be seamlessly integrated into physics-based NMAR frameworks to simplify image reconstruction pipelines and enhance computational performance. The proposed unified forward projection provides an efficient solution to accelerate MAR in CT imaging.

To evaluate whether a dedicated, anatomically realistic knee phantom with orthopedic implants can be used to optimize clinical MRI sequences under realistic metallic conditions and to assess whether phantom-optimized sequences improve image quality compared with routine protocols. A 3D-printed knee phantom integrating a titanium screw and a stainless-steel fixation plate was developed. Phantom imaging was performed on a clinical 1.5T MRI system. In this exploratory proof-of-concept study, T1w and STIR sequences were iteratively optimized using the phantom and compared with vendor-default and routine clinical protocols. The phantom-optimized sequences were subsequently applied in three asymptomatic volunteers with metallic knee implants and compared with routine sequences. Image quality was independently assessed by four blinded readers using a 5-point Likert scale across spatial resolution, artifact reduction, and overall image quality. Phantom-optimized sequences achieved significantly higher image quality scores compared with routine protocols. Improvements were consistent across all three predefined criteria: spatial resolution (mean increase +0.70 points, p < 0.001), artifact reduction (+0.65, p < 0.001), and overall image quality (+0.78, p < 0.001). These gains were observed for both T1-weighted and STIR acquisitions without extending acquisition times. Phantom-guided optimization provides a reproducible, patient-independent framework for tuning MRI protocols near orthopedic hardware. Anatomically realistic phantoms represent a promising methodological tool for developing, testing, and standardizing MRI sequences under controlled and clinically relevant conditions.

BackgroundThe development of PET scanners dedicated to high temporal and spatial resolution organ‐specific imaging is an active research area, motivated by the need for cost reduction, improved lesion detectability and quantification in specific clinical scenarios, as well as by ongoing hardware and software innovations.PurposeThis study investigates and compares various image reconstruction strategies for a dual‐panel prostate‐dedicated PET scanner (ProVision), which features four‐layered dual‐readout time‐of‐flight depth‐of‐interaction detectors and a 22‐position acquisition protocol to improve angular coverage.Materials and methodsA list‐mode MLEM algorithm with multi‐ray modeling was developed and optimized using a scaled NEMA image quality phantom to determine optimal number of rays and iterations. These parameters were then used to reconstruct data from both simulation and experimental acquisitions, including an anthropomorphic pelvis phantom, named Adam‐PETer. Four reconstruction approaches were evaluated: classical MLEM; MLEM with embedded shift‐variant point spread function (PSF) modeling; a hybrid list‐mode reconstruction; and a Swin‐UNETR‐based deep learning model applied as a post‐reconstruction enhancement to MLEM images. Performance was assessed using contrast recovery coefficient (CRC), contrast‐to‐noise ratio (CNR), and contrast‐to‐noise consistency (CNC), on both a scaled NEMA phantom and an experimental anthropomorphic phantom.ResultsIn the scaled NEMA phantom simulation, the Swin‐based method yielded the highest CNR and CNC, especially for the smallest spheres, thereby outperforming both standard MLEM and the hybrid algorithm. In the Adam‐PETer experimental prostate phantom, the CNR was 10.43 for MLEM, 14.48 for Hybrid, and 13.85 for Swin for the larger lesion (10 mm). The CNR values were 2.28, 3.03, and 4.35, respectively, for the smaller lesion (8 mm). CNC values also varied across methods, with Swin achieving the best result for the smaller lesion. These findings indicate that model‐based and learned methods offer complementary strengths depending on lesion contrast and size.ConclusionThe PET scanner‐adapted reconstruction combined with deep learning refinement improves image quality in dedicated, limited‐angle PET systems.

Arthrofibrosis is a common complication following total knee arthroplasty (TKA), characterized by excessive fibrous tissue formation within the joint, leading to restricted range of motion (ROM), pain, and functional impairment. Accurate diagnosis is essential for distinguishing arthrofibrosis from other causes of postoperative knee stiffness, such as infection, mechanical block, or malalignment. This review aims to explore current diagnostic methods and evolving standards for arthrofibrosis after TKA, focusing on (1) clinical differentiation from other causes of knee stiffness; (2) assessment and diagnostic criteria; (3) imaging, laboratory, and histopathological techniques; and (4) an integrated diagnostic algorithm and future directions. Diagnosis is primarily based on persistent ROM limitation (flexion <90 degrees or extension >5 degrees) for more than 12 weeks, after excluding infection and mechanical causes. Advanced magnetic resonance imaging (MRI) with metal artifact reduction techniques can be used to visualize intra-articular fibrosis, with an MRI-based synovial classification correlating with ROM deficits and severity. Synovial fluid analysis helps rule out infection, and histopathology is employed when the diagnosis remains unclear. The study proposes a stepwise diagnostic algorithm that integrates clinical, imaging, and laboratory findings and discusses future directions for optimizing diagnosis and treatment pathways to improve patient outcomes.

A modified interleaved flyback (miFB) approach is introduced, designed to mitigate flow artifacts caused by alternating readout polarities in Echo Planar Imaging (EPI), while preserving acquisition efficiency. We propose reconstructing odd and even echoes of 3D-EPI separately. To this end, the respective missing lines are acquired in interleaved shots with inverted polarity and an additional gradient pre-lobe. Thereby, high scan efficiency is maintained compared to unsampled flyback gradients. Our miFB approach is additionally combined with gradient moment smoothing and compared to the interleaved dual-echo with acceleration (IDEA) method in phantom and in vivo scans at 7 Tesla. Our results demonstrate a significant reduction in ghosting and signal dropout using the miFB approach, yielding comparable image quality to non-EPI acquisitions while reducing acquisition time by approximately half. The miFB approach offers a substantial reduction in flow artifacts, allowing for decreased acquisition times in TOF-MRA.

BackgroundBeam hardening is an unavoidable phenomenon in polychromatic CT systems, where lower‐energy photons are preferentially absorbed as x‐rays pass through materials such as tissue, bone, and metal. This energy‐dependent attenuation, which conflicts with the monochromatic assumption in filtered back‐projection (FBP), produces artifacts that degrade CT image quality and diagnostic accuracy.PurposeThis study aims to quantitatively estimate the mean energy corresponding to the beam‐hardened projection data along each x‐ray path and to employ this mean energy to convert energy‐dependent projections into beam‐hardening–corrected data. The effectiveness of the proposed mean energy–based correction method is verified through numerical simulations.MethodsTo mathematically determine the mean energy, a polynomial equation is derived whose solution represents the mean energy. Based on the Beer–Lambert law, the CT system is formulated as an energy‐integrated model incorporating the x‐ray spectrum and the linear attenuation coefficient of the scanned object. By applying the mean value theorem for integrals and performing power series expansions of the energy‐dependent components of the attenuation coefficient—such as Compton scattering and photoelectric absorption—a polynomial equation with respect to the mean energy is obtained. Solving this equation yields the mean energy, which is subsequently employed to generate beam‐hardening—corrected projection data.ResultsA novel correction method based on the computed mean energy is proposed to transform energy‐dependent projection data into corresponding monochromatic data. Numerical simulations conducted on various models demonstrate that the proposed approach effectively corrects beam‐hardened projection data and substantially reduces beam‐hardening artifacts in reconstructed CT images. The method maintains accuracy even in complex scenarios involving multiple overlapping materials and metallic objects, without a significant increase in computational cost. Furthermore, the robustness of the proposed correction technique is confirmed under varying x‐ray spectra, verifying its potential applicability to practical CT imaging.ConclusionsGrounded in physical modeling and analytical approximation, this study presents a mathematical formulation for estimating the mean energy corresponding to beam‐hardened projection data and develops a correction method that effectively mitigates beam‐hardening artifacts. The results highlight the potential of the proposed mean energy—based correction as a practical and computationally efficient solution for improving CT image quality. However, as this study primarily focused on the computation of mean energy as an indicator of beam‐hardening severity, further research is required to apply and validate the proposed method using experimental and clinical data for comprehensive verification of its practical applicability.

Brain magnetic resonance imaging (MRI) is highly susceptible to motion artifacts that degrade fine structural details and undermine quantitative analysis. Conventional U-Net-based deep learning approaches for motion artifact reduction typically operate only in the image domain and are often trained on data with simplified motion patterns, thereby limiting physical plausibility and generalization. We propose Sim-DDNet, a simulation-data-based dual-domain network that combines k-space-based motion simulation with a joint image-k-space reconstruction architecture. Motion-corrupted data were generated from T2-weighted Alzheimer’s Disease Neuroimaging Initiative brain MR scans using a k-space replacement scheme with three to five random rotational and translational events per volume, yielding 69,283 paired samples (49,852/6969/12,462 for training/validation/testing). Sim-DDNet integrates a real-valued U-Net-like image branch and a complex-valued k-space branch using cross attention, FiLM-based feature modulation, soft data consistency, and composite loss comprising L1, structural similarity index measure (SSIM), perceptual, and k-space-weighted terms. On the independent test set, Sim-DDNet achieved a peak signal-to-noise ratio of 31.05 dB, SSIM of 0.85, and gradient magnitude similarity deviation of 0.077, consistently outperforming U-Net and U-Net++ across all three metrics while producing less blurring, fewer residual ghost/streak artifacts, and reduced hallucination of non-existent structures. These results indicate that dual-domain, data-consistency-aware learning, which explicitly exploits k-space information, is a promising approach for physically plausible motion artifact correction in brain MRI.

Electroencephalography (EEG) data can be used to decode an attended sound source in normal-hearing (NH) listeners, even for music stimuli. This information could steer the sound processing strategy for cochlear implants (CIs) users, potentially improving their music listening experience. The aim of this study was to investigate whether selective auditory attention decoding (SAAD) could be performed in CI users for music stimuli.&#xD;Approach: High-density EEG was recorded from 8 NH and 8 CI users. Duets containing a clarinet and cello were dichotically presented. A linear decoder was trained to reconstruct audio features of the attended instrument from EEG data. The estimated attended instrument was selected based on which of the two instruments had a higher correlation to the reconstructed instrument. EEG recordings are challenging in CI users, as these devices introduce strong electrical artifacts. We also propose a new artifact rejection technique that employs ICA calculating ICs and automating their selection for removal, which we termed ASICA.&#xD;Main results: &#xD;We showed that it was possible to perform SAAD for music in CI users. The decoding accuracies were 59.4 \% for NH listeners and 60 \% for CI users with the proposed algorithm. &#xD;Using the proposed algorithm, the correlation coefficients between the reconstructed audio feature and the attended audio feature were improved in conditions where artifact was dominating. &#xD;Significance: &#xD;Results indicate that selective auditory attention to musical instruments can be effectively decoded, and that this decoding is enhanced by the new artifact reduction algorithm, particularly in scenarios where the cochlear implant's electrical artifact has greater influence.&#xD;Moreover, these results could be relevant as an objective measure of music perception or for a brain computer interface that improves music enjoyment. Additionally we showed that the stimulation artifact can be suppressed. &#xD;The ethic's committee of the MHH approved this study (8874_BO_K_2020).

Double inversion recovery (DIR) is an MRI technique in which two types of tissue are suppressed, usually cerebrospinal fluid (CSF) and white matter (WM). The suppression is achieved with two inversion pulses prior to the acquisition of the imaging data. In the presence of strong inhomogeneities in the static magnetic field B0 and/or the radiofrequency (RF) field, inversion may be inadequate, resulting in bright signal in tissues that should have been suppressed. The purpose of this work was to develop a DIR scan with inversion pulses that are robust against inhomogeneities in the B0 and RF field. In this prospective study, the DIR sequence was equipped with inversion pulses designed with optimal control. Robustness against field inhomogeneities was incorporated into the cost functional for pulse optimization. DIR and controlled DIR (C-DIR) MRI images were acquired at 3T in 14 participants (9 male, age=36.1±11.5 years) enrolled between October 2024 to August 2025 from a single academic medical center: nine healthy; two with relapsing-remitting multiple sclerosis; one with persistent concussion symptoms; two with asymptomatic white matter hyperintensities. Suppression of CSF, presence of artifacts, and visibility of multiple sclerosis lesions and white matter hyperintensities were independently assessed visually by a radiologist. In eight healthy volunteers, means and SDs were computed for signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), with significance evaluated using a Student's t-test. C-DIR exhibits improved inversion in the presence of inhomogeneities in the B0 and the radiofrequency field, resulting in the removal of artifactual signal. CNR increases ranged from 27% between gray matter and CSF (p<.001) to 102% between the brainstem and adjacent CSF (p<0.001). SNR in the cortical gray matter was 10.74±1.48 in DIR and 11.68±2.21 in C-DIR (p=.07). Inversion with a robust RF pulse improves the quality of DIR, demonstrating artifact reduction and improved CNR.

Despite the increasing use of MARS (metal artifact reduction sequence) MRI to investigate painful total hip arthroplasties (THA), no validated criteria exist for diagnosing femoral stem loosening. To evaluate MARS MRI for the diagnosis of aseptic stem loosening and determine its diagnostic accuracy. One hundred fourteen consecutive cases with THA revision surgery and MARS MRI of the hip were retrospectively included. Two blinded musculoskeletal radiologists independently assessed periprosthetic bone resorption (PPBR), bone marrow edema (BME), periosteal reaction (PR), and periprosthetic osteolysis (PO) in 14 Gruen zones (GZ). Intraoperative findings at revision surgery served as the ground truth. A predictive model was created using binomial logistic regression models to predict the probability of a loose stem with maximizing positive predictive value (PPV) and accuracy. Interobserver reliability was assessed with absolute agreement, Cohen κ and Gwet AC1. During surgery, 66 stems were fixed and 48 loose. PPBR occurred significantly more frequently in loose stems across all GZs except GZ11. Proximal PPBR was also observed in fixed stems (up to 23%), whereas middle and distal PPBR were rare (≤3%). BME was most prevalent proximally in all stems (fixed/loose: 39%/60%) with significant differences medially and distally. PR was significantly more frequent in loose stems in the middle and distal GZs. PO were rare, most occurred in GZ7. The predictive model considering proximal PPBR, mid-distal PPBR, mid-distal PR, and distal BME performed with a sensitivity of 0.708, specificity of 0.970, PPV 0.944, negative predictive value 0.821. Interobserver agreement (Gwet AC1) in the considered zones was for PPBR between 0.80 and 0.98, BME 0.91 to 0.99, PR 0.87 to 0.97. MARS MRI is reproducible and accurate for assessing stem loosening. PPBR, BME, and PR can also be found in fixed THA in the proximal region, whereas they indicate loosening in the middle and distal stem region.

PurposeTo present a non-mydriatic hyperspectral retinal camera based on spectral scanning, developed to achieve a practical balance among imaging performance, acquisition speed, and system simplicity for advanced retinal diagnostics.MethodsThe system integrates LED-based broadband illumination, linear variable filters, custom optics, a monochrome sensor, and a motorized three-dimensional stage to capture high-resolution hyperspectral data across 29 wavebands from 450 to 850 nm. Optical performance was evaluated using standard metrics including spectral resolution, irradiance, uniformity, resolving power, field of view, and chromatic focal compensation. Imaging was performed on model eyes and human subjects to assess spectral signature capture and repeatability.ResultsThe system achieved a 40° field of view and a spectral resolution ranging from 20 to 80 nm. Chromatic focal correction and illumination uniformity were maintained across the spectral range. In vivo imaging demonstrated the ability to capture distinct spectral signatures of anatomical structures and ocular pathologies. Test–retest assessments showed high repeatability, with spectral variation below 5%. The device operated under non-mydriatic conditions with acquisition times of approximately 300 ms.ConclusionsThe prototype demonstrates reliable and repeatable hyperspectral imaging of the retina in a compact and semi-automated form factor. The system offers a foundation for further optimization, including improved spectral precision, artifact reduction, and increased field of view.Translational RelevanceThis technology enables non-invasive, high-content retinal imaging suitable for integration into clinical workflows.

The role of AI in optimizing CMR image quality: A scoping review.

Deep learning motion correction of quantitative stress perfusion cardiovascular magnetic resonance.

Truncated projection adaptive weighting combined with adaptive TV for artifact reduction in linear computed laminography