Separate Maps Research Articles

Self-supervised parametric map estimation for multiplexed PET with a deep image prior.

Multiplexed positron emission tomography (mPET) imaging allows simultaneous observation of physiological and pathological information from multiple tracers in a single PET scan. Although supervised deep learning has demonstrated superior performance in mPET image separation compared to purely model-based methods, acquiring large amounts of paired single-tracer data and multi-tracer data for training poses a practical challenge and needs extended scan durations for patients. In addition, the generalisation ability of the supervised learning framework is a concern, as the patient being scanned and their tracer kinetics may potentially fall outside the training distribution. In this work, we propose a self-supervised learning framework based on the deep image prior (DIP) for mPET image separation using just one dataset. In particular, we integrate the multi-tracer compartmental model into the DIP framework to estimate the parametric maps of each tracer from the measured dynamic dual-tracer activity images. Consequently, the separated dynamic single tracer activity images can be recovered from the estimated tracer-specific parametric maps. In the proposed method, dynamic dual-tracer activity images are used as the training label, and the static dual-tracer image (reconstructed from the same patient data from the start to the end of acquisition) is used as the network input. The performance of the proposed method was evaluated on a simulated brain phantom for dynamic dual-tracer [18F]FDG+[11C]MET activity image separation and parametric map estimation. The results demonstrate that the proposed method outperforms the conventional voxel-wise multi-tracer compartmental modeling method (vMTCM) and the two step method DIP-Dn+vMTCM (where dynamic dual-tracer activity images are first denoised using a U-net within the DIP framework, followed by vMTCM separation) in terms of lower bias and standard deviation in the separated single-tracer images and also for the estimated parametric maps for each tracer, at both voxel and ROI levels.

Evaluation of environmental factors related to Aspergillus fumigatus azole resistance in the Netherlands

Emergence of azole-resistant Aspergillus fumigatus represents a concern for public health worldwide. Here we applied a spatial analysis to a large number of records on A. fumigatus azole resistance in the Netherlands to evaluate the relationship of a set of environmental factors with the emergence of resistant isolates and to identify the potential risk areas. A total of 1850 aspergillosis cases were included in the analysis: 1559 caused by wild-type (WT) and 291 by azole-resistant (RES) strains. High-resolution maps containing environmental variables (climate, crop cultivations) were used as layers for spatial analysis. QGIS software was used for transformation and calculation of layers, and visualizing maps, whereas MaxEnt software was used to perform spatial analysis. A separate distribution map for WT and RES was generated using each layer, then the ratio between RES and WT suitability was calculated, and a new map was generated showing the areas where suitability of RES was higher than WT (ratio > 1). Layers presenting areas with a ratio > 1 were then used for a further analysis including all these layers together to generate a final risk map, to identify the most contributing environmental variables, and to estimate the number of people potentially exposed to the risk. Results showed that the areas with the highest RES/WT ratio were associated with the following variables: legume cultivations ratio 1.5–2.7; fruit tree plantations, ratio 2.1; dump sites and artificial sites, ratio 2.3–2.7; spring minimum and mean temperatures, ratio 1.6–1.8. Areas with medium, high, and very high risk covered a surface of about 2330 km2 and people exposed were >319,000, with Zeeland and Noord-Holland being the provinces with the highest risk. All together these results suggest that azole-resistance emergence is a complex phenomenon depending on a multitude of environmental variables that are not yet been fully explored.

Subcellular Proteomic Mapping of Lysine Lactylation.

Protein lactylation is a novel post-translational modification (PTM) involved in many important physiological processes such as macrophage polarization, immune regulation, and tumor cell growth. However, traditional methodologies for studying lactylation have predominantly relied on peptide enrichment from whole-cell lysates, which tend to favor the detection of high-abundance peptides, thus limiting the identification of low-abundance lactylated peptides. To address this limitation, here, we employed subcellular fractionation to separate proteins and map lactylated peptides from each isolated subcellular fraction using a model cell line. In brief, we identified 1,217 lysine lactylation (Kla) sites on 553 proteins across four subcellular fractions. Subsequent pathway enrichment analysis revealed that Kla proteins participate in distinct pathways depending on the subcellular contexts. In addition, this subcellular fractionation method enabled the discovery of 36 previously unreported Kla proteins and 223 novel Kla sites, many of which are present in low abundance. Notably, several proteins contain multiple newly identified Kla sites, exemplified by the transcriptional regulator ATRX. Furthermore, our results indicate the possibility of PTM crosstalk between Kla and other PTMs such as ubiquitination and sumoylation. In conclusion, subcellular fractionation facilitates the identification of Kla proteins that have been previously uncovered and could be overlooked by affinity enrichment of whole-cell lysates.

3D B1+ corrected simultaneous myocardial T1 and T1ρ mapping with subject-specific respiratory motion correction and water-fat separation.

To develop a 3D free-breathing cardiac multi-parametric mapping framework that is robust to confounders of respiratory motion, fat, and B1+ inhomogeneities and validate it for joint myocardial T1 and T1ρ mapping at 3T. An electrocardiogram-triggered sequence with dual-echo Dixon readout was developed, where nine cardiac cycles were repeatedly acquired with inversion recovery and T1ρ preparation pulses for T1 and T1ρ sensitization. A subject-specific respiratory motion model relating the 1D diaphragmatic navigator to the respiration-induced 3D translational motion of the heart was constructed followed by respiratory motion binning and intra-bin 3D translational and inter-bin non-rigid motion correction. Spin history B1+ inhomogeneities were corrected with optimized dual flip angle strategy. After water-fat separation, the water images were matched to the simulated dictionary for T1 and T1ρ quantification. Phantoms and 10 heathy subjects were imaged to validate the proposed technique. The proposed technique achieved strong correlation (T1: R2 = 0.99; T1ρ: R2 = 0.98) with the reference measurements in phantoms. 3D cardiac T1 and T1ρ maps with spatial resolution of 2 × 2 × 4 mm were obtained with scan time of 5.4 ± 0.5 min, demonstrating comparable T1 (1236 ± 59 ms) and T1ρ (50.2 ± 2.4 ms) measurements to 2D separate breath-hold mapping techniques. The estimated B1+ maps showed spatial variations across the left ventricle with the septal and inferior regions being 10%-25% lower than the anterior and septal regions. The proposed technique achieved efficient 3D joint myocardial T1 and T1ρ mapping at 3T with respiratory motion correction, spin history B1+ correction and water-fat separation.

Organizational culture and leadership style in Spanish Hospitals: Effects on knowledge management and efficiency

Hospital culture and leadership style have attracted considerable attention in research, with compelling evidence indicating their potential competitive advantages, including their crucial role in ensuring the successful implementation of knowledge management and its impact on hospital efficiency. The aim of this paper is to identify the effects of organizational culture and leadership style on knowledge management and hospital efficiency. Fuzzy cognitive maps (FCMs) are relational models that can be used to represent the opinions and knowledge of expert to infer cause-effect relationships among different concepts. The use of FCMs as a simulation tool enables the evaluation of potential scenarios based on different organizational cultures and leadership styles in hospitals. Developing an FCM for this study involved several steps. Firstly, data were collected through interviews with 21 experts in hospital management. The interviews were conducted between May and September 2023 either face-to-face or via videoconference. Once individual cognitive maps had been created, consensus among them was achieved through a multicriteria decision-making process, wherein the expert opinions were averaged. The separate cognitive maps of each expert were then integrated to produce a single FCM using the augmented FCM approach. Reflecting expert insights from the FCM, hospitals with a hierarchy culture exhibit diminished levels of knowledge creation, management, and overall hospital efficiency, whereas those with an adhocracy culture show improvements in knowledge creation, knowledge exploitation, and overall hospital efficiency in comparison to alternative ones. From the experts' FCM perspective regarding leadership style, transformational leadership achieves the highest level of knowledge management and hospital efficiency in hospitals with an adhocracy culture. Finally, this paper offers a reference for practising knowledge management and improving hospital efficiency through adhocracy culture and transformational leadership.

Accurate crowd counting for intelligent video surveillance systems

The paper presents a novel deep learning approach for crowd counting in intelligent video surveillance systems, addressing the growing need for accurate monitoring of public spaces in urban environments. The demand for precise crowd estimation arises from challenges related to security, public safety, and efficiency in urban areas, particularly during large public events. Existing crowd counting techniques, including feature-based object detection and regression-based methods, face limitations in high-density environments due to occlusions, lighting variations, and diverse human figures. To overcome these challenges, the authors propose a new deep encoder-decoder architecture based on VGG16, which incorporates hierarchical feature extraction with spatial and channel attention mechanisms. This architecture enhances the model’s ability to manage variations in crowd density, leveraging adaptive pooling and dilated convolutions to extract meaningful features from dense crowds. The model’s decoder is further refined to handle sparse and crowded scenes through separate density maps, improving its adaptability and accuracy. Evaluations of the proposed model on benchmark datasets, including Shanghai Tech and UCF CC 50, demonstrate superior performance over state-of-the-art methods, with significant improvements in mean absolute error and mean squared error metrics. The paper emphasizes the importance of addressing environmental variability and scale differences in crowded environments and shows that the proposed model is effective in both sparse and dense crowd conditions. This research contributes to the advancement of intelligent video surveillance systems by providing a more accurate and efficient method for crowd counting, with potential applications in public safety, transportation management, and urban planning.

Interoperability Verification of Wireless Power Transfer System with Separate Reference Impedance Map Method

Interoperability Verification of Wireless Power Transfer System with Separate Reference Impedance Map Method

Optimised path planning using Enhanced Firefly Algorithm for a mobile robot.

Path planning is a crucial element of mobile robotics applications, attracting considerable interest from academics. This paper presents a path-planning approach that utilises the Enhanced Firefly Algorithm (EFA), a new meta-heuristic technique. The Enhanced Firefly Algorithm (FA) differs from the ordinary FA by incorporating a linear reduction in the α parameter. This modification successfully resolves the constraints of the normal FA. The research involves experiments on three separate maps, using the regular FA and the suggested Enhanced FA in 20 different runs for each map. The evaluation criteria encompass the algorithms' ability to move from the initial location to the final position without experiencing any collisions. The assessment of path quality relies on elements such as the distance of the path and the algorithms' ability to converge and discover optimum solutions. The results demonstrate significant improvements made by the Enhanced FA, with a 10.270% increase in the shortest collision-free path for Map 1, a 0.371% increase for Map 2, and a 0.163% increase for Map 3, compared to the regular FA. This work highlights the effectiveness of the Enhanced Firefly Algorithm in optimising path planning for mobile robotics applications, providing potential improvements in navigation efficiency and collision avoidance.

Encoding of 2D Self-Centered Plans and World-Centered Positions in the Rat Frontal Orienting Field.

The neural mechanisms of motor planning have been extensively studied in rodents. Preparatory activity in the frontal cortex predicts upcoming choice, but limitations of typical tasks have made it challenging to determine whether the spatial information is in a self-centered direction reference frame or a world-centered position reference frame. Here, we trained male rats to make delayed visually guided orienting movements to six different directions, with four different target positions for each direction, which allowed us to disentangle direction versus position tuning in neural activity. We recorded single unit activity from the rat frontal orienting field (FOF) in the secondary motor cortex, a region involved in planning orienting movements. Population analyses revealed that the FOF encodes two separate 2D maps of space. First, a 2D map of the planned and ongoing movement in a self-centered direction reference frame. Second, a 2D map of the animal's current position on the port wall in a world-centered reference frame. Thus, preparatory activity in the FOF represents self-centered upcoming movement directions, but FOF neurons multiplex both self- and world-reference frame variables at the level of single neurons. Neural network model comparison supports the view that despite the presence of world-centered representations, the FOF receives the target information as self-centered input and generates self-centered planning signals.

Computationally Hard Problems for Logic Programs under Answer Set Semantics

Showing that a problem is hard for a model of computation is one of the most challenging tasks in theoretical computer science, logic and mathematics. For example, it remains beyond reach to find an explicit problem that cannot be computed by polynomial size propositional formulas (PF). As a model of computation, logic programs (LP) under answer set semantics are as expressive as PF, and also \(\mathtt{NP}\) -complete for satisfiability checking. In this paper, we show that the PAR problem is hard for LP, i.e., deciding whether a binary string contains an odd number of \(1\) ’s requires exponential size logic programs. The proof idea is first to transform logic programs into equivalent boolean circuits, and then apply a probabilistic method known as random restriction to obtain an exponential lower bound. Based on the main result, we generalize a sufficient condition for identifying hard problems for LP, and give a separation map for a logic program family from a computational point of view, whose members are all equally expressive and share the same reasoning complexity.

Serial and parallel organ-at-risk-specific noncoplanar arc optimization for small versus large target volumes in liver SBRT.

Noncoplanar arc optimization has been shown to reduce OAR doses in SRS/SRT and has the potential to reduce doses to OARs in SBRT. Extracranial targets have additional considerations, including large OARs and, in the case of the liver, volume constraints on the healthy liver. Considering pathlengths through OARs that encompass target volumes may lead to specific dose reductions as in the encompassing healthy liver tissue. These optimizations must also leverage delivery efficiency and trajectory sampling to ensure ease of clinical translation. The purpose of this research is to generate optimized static-couch arcs that separately consider serial and parallel OARs and arc delivery efficiency, with a trajectory sampling metric, towards the aim of reducing dose to OARs and the surrounding healthy liver tissue. Separate BEV cost maps were created for parallel, and serial OARs by means of a fast ray-triangle intersection algorithm. An additional BEV cost map was created for the liver which, by definition, encompasses the liver tumors. The individual costs of these maps were summed and combined with the sampling metric for 100 000 random combinations of arc trajectories. A search algorithm was applied to find an arc trajectory solution that satisfied BEV cost and sampling optimization, while also ensuring an efficient delivery was possible with a low number of arcs. This method of arc selection was evaluated for 16 liver SBRT patients characterized by small and large target volumes. Comparisons were made with a clinical arc template of coplanar arcs. Dosimetric plan quality was evaluated using published guidelines and metrics from RTOG1112. Four of five plan quality metrics for the liver were significantly reduced when planned with optimized noncoplanar arcs. Median (range) reductions of the volumes receiving 10, 18, and 21Gy were found of 140.4 (295.8) cc (p=0.001), 28.2 (230.6) cc (p=0.002) and 18.5 (155.5) cc (p=0.04). A significant increase in median (range) dose to the right kidney of 0.2±0.9Gy (p=0.03) was also found using optimized noncoplanar arcs, which was below the tolerance of 10Gy for all cases. The average number of arcs chosen was 4±1. Optimizing serial and parallel OARs separately during static couch noncoplanar arc selection significantly reduced the dose to the liver during SBRT using a moderate number of arcs.

High-precision polarization imaging for three-dimensional reconstruction aided by a separate coarse depth map

For less-texture objects with highly reflective regions, traditional vision-based three-dimensional reconstruction techniques often fail to yield ideal results. Utilizing polarization information for reconstructing such objects is a convenient and effective method. However, relying solely on polarization information for three-dimensional reconstruction presents challenges such as ambiguity in surface normals and difficulties in normal integration. In this paper, we propose to resolve the ambiguity in polarization normals using the coarse depth as prior information under a perspective projection model. By fusing the disambiguated normal with the coarse depth, we avoid the need for normal integration. The experimental results demonstrate that our proposed algorithm improves the quality of polarization imaging, effectively restoring the details lost in the coarse depth and smoothing the areas with high reflectivity.

A new lightweight network for efficient UAV object detection

Optimizing the structure of deep neural networks is essential in many applications. Especially in the object detection tasks of Unmanned Aerial Vehicles. Due to the constraints of the onboard platform, a more efficient network is required to meet practical demands. Nevertheless, existing lightweight detection networks exhibit excessive redundant computations and may yield in a certain level of accuracy loss. To address these issues, this paper proposes a new lightweight network structure named Cross-Stage Partially Deformable Network (CSPDNet). The initial proposal consists of a Deformable Separable Convolution Block (DSCBlock), separating feature channels, greatly reducing the computational load of convolution, and applying adaptive sampling to the separated feature map. Subsequently, to establish information interaction between feature layers, a channel weighting module is proposed. This module calculates weights for the separated feature map, facilitating information exchange across channels and resolutions. Moreover, it compensates for the effect of point-wise (1 ×\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ imes$$\\end{document} 1) convolutions, filtering out more important feature information. Furthermore, a new CSPDBlock is designed, primarily composed of DSCBlock, establishing multidimensional feature correlations for each separated feature layer. This approach improves the ability to capture critical feature information and reconstruct gradient paths, thereby preserving detection accuracy. The proposed technology achieves a balance between model parameter size and detection accuracy. Furthermore, experimental results on object detection datasets demonstrate that our designed network, using fewer parameters, achieves competitive detection performance results compared to existing lightweight networks YOLOv5n, YOLOv6n, YOLOv8n, NanoDet and PP-PicoDet. The optimization effect of the designed CSPDBlock, using the VisDrone dataset, is validated when incorporated into advanced detection algorithms YOLOv5m, PPYOLOEm, YOLOv7, RTMDetm and YOLOv8m. In more detail, by incorporating the designed modules it was achieved that the parameters were reduced by 10–20% while almost maintaining detection accuracy.

H-SLAM: Hybrid direct–indirect visual SLAM

The recent success of hybrid methods in monocular odometry has led to many attempts to generalize the performance gains to hybrid monocular SLAM. However, most attempts fall short in several respects, with the most prominent issue being the need for two different map representations (local and global maps), with each requiring different, computationally expensive, and often redundant processes to maintain. Moreover, these maps tend to drift with respect to each other, resulting in contradicting pose and scene estimates, and leading to catastrophic failure. In this paper, we propose a novel approach that makes use of descriptor sharing to generate a single inverse depth scene representation. This representation can be used locally, queried globally to perform loop closure, and has the ability to re-activate previously observed map points after redundant points are marginalized from the local map, eliminating the need for separate map maintenance processes. The maps generated by our method exhibit no drift between each other, and can be computed at a fraction of the computational cost and memory footprint required by other monocular SLAM systems. Despite the reduced resource requirements, the proposed approach maintains its robustness and accuracy, delivering performance comparable to state-of-the-art SLAM methods (e.g., LDSO, ORB-SLAM3) on the majority of sequences from well-known datasets like EuRoC, KITTI, and TUM VI. The source code is available at: https://github.com/AUBVRL/fslam_ros_docker.

Tracking and co-location of global point clouds for large-area indoor environments

Extended reality (XR) experiences are on the verge of becoming widely adopted in diverse application domains. An essential part of the technology is accurate tracking and localization of the headset to create an immersive experience. A subset of the applications require perfect co-location between the real and the virtual world, where virtual objects are aligned with real-world counterparts. Current headsets support co-location for small areas, but suffer from drift when scaling up to larger ones such as buildings or factories. This paper proposes tools and solutions for this challenge by splitting up the simultaneous localization and mapping (SLAM) into separate mapping and localization stages. In the pre-processing stage, a feature map is built for the entire tracking area. A global optimizer is applied to correct the deformations caused by drift, guided by a sparse set of ground truth markers in the point cloud of a laser scan. Optionally, further refinement is applied by matching features between the ground truth keyframe images and their rendered-out SLAM estimates of the point cloud. In the second, real-time stage, the rectified feature map is used to perform localization and sensor fusion between the global tracking and the headset. The results show that the approach achieves robust co-location between the virtual and the real 3D environment for large and complex tracking environments.

Visualization of Solar Radiation Using Three-Dimensional Computer Graphics Technologies

Currently, the prediction, calculation and rationing of solar radiation is a serious problem that has a wide impact on the spheres of human activity. Creating and maintaining a comfortable living environment, as well as solving the world's energy problems, are the fundamental criteria that determine its relevance. This article discusses the possibility of using three-dimensional graphics technologies to calculate insolation. The paper considers an algorithm for calculating the four main components of the visualization of global solar radiation: direct beam radiation, diffuse radiation, reflected beam radiation and reflected diffuse radiation. The freely distributed open source software Blender 3D is used as a tool for working with computer graphics. In the course of the study, calculations based on the principle of bidirectional path tracing, which is used in the visualization algorithms of the Cycles graphics engine, are presented. The use of the technology of transferring complex data to the raster image format allows you to create a separate texture map for each component of the simulated lighting. Based on the obtained texture maps, it is possible to calculate the global solar radiation for all created or imported 3D models in the three-dimensional space of Blender 3D.

Improved skeleton-based activity recognition using convolutional block attention module

Inferring human activities from the skeletons extracted from activity photos or videos is a fundamental yet important issue in the research community of computer vision. Current skeleton-based activity recognition methods have accomplished this task by extracting features in either spatial or temporal perspective, or combining both views; however, the extracted features lack further after-processing like feature augmentation. Aiming at the above shortcomings, this article presents a new human activity recognition approach based on human skeleton data, named Convolutional Block Attention Module-based Spatial Temporal Convolutional Network (CBAM-STGCN). To enhance the discriminative validity of the generated features, we leverage the attention mechanism to weight feature units within each dimension by introducing two attention modules designed in terms of different aspects into the model architecture. Specifically, the original feature map of an activity sample is adaptively refined by the convolutional block attention module, where two separate attention maps obtained with regard to spatial and channel dimensions are adopted to generate the corresponding attended feature maps. Evaluation results on the Kinetics benchmark dataset showed that the proposed model has improved 1.76 % and 2.43 % according to top-1 and top-5 metrics, respectively, compared to the conventional baseline model. Further ablation experiments proved that both channel and spatial attention modules contribute to the model performance.

Abstract 4948: Automated tumor microenvironment analysis for multiple samples by image-based spatial transcriptomics on tissue microarray

Abstract Introduction. Tissue Microarrays (TMAs), widely utilized in the field of pathology, have now found a powerful ally in Image-based Spatial Transcriptomics (ST). By analyzing various gene expression data with high resolution, image-based ST data on TMA can provide the heterogeneous patterns of tumor microenvironment in multiple samples. However, an efficient method for processing multiple samples post-data acquisition is still under development. A streamlined process would expedite the discovery of spatial gene expression patterns, thereby enhancing our understanding of the tumor microenvironment and its implications for cancer diagnostics and treatment strategies. Methods. Our automated pipeline is initiated by automatic segmentation of every core in the whole TMA, derived from the processed output of the MERSCOPE or Xenium platform. Then, it subsequently removes QC failed cells unrelated to any core. Each core receives sequential naming and undergoes automated cell type mapping utilizing a reference single-cell RNAseq data. Additionally, the pipeline computes neighborhood enrichment between cell types, providing a nuanced comprehension of spatial relationships and interactions among diverse cell populations within the tissue microenvironment from multiple samples on the TMA. Results. The automated pipeline we've developed provides several key advantages. It enables the simultaneous analysis of multiple tissue cores, effectively minimizing batch effects and ensuring the reliability of results across diverse tissue samples. Furthermore, by automating core separation, labeling, and cell type mapping, our pipeline significantly streamlines the time and effort required for TMA data analysis. This cost-effective approach allows researchers to optimize resource allocation, making our automated pipeline a valuable tool in cancer research. It facilitates the exploration of gene expression patterns within specific tissue regions, ultimately contributing to the advancement of our understanding of cancer biology. Conclusion. By seamlessly integrating TMA data with image-based ST technologies such as Xenium and MERSCOPE, we can perform comprehensive spatial transcriptomics analyses and provide detailed statistical information on cell type proportions within the tumor microenvironment. This automated pipeline not only ensures robust and reliable results across diverse tissue samples but also optimizes resource allocation by minimizing batch effects and reducing analysis time. This cost-effective solution empowers researchers to delve deeper into spatial gene expression patterns, enhancing our comprehension of the tumor microenvironment. Citation Format: Dongjoo Lee, Seungho Cook, Yeonjae Jung, Myunghyun Lim, Jae Eun Lee, Hyung-Jun Im, Daeseung Lee, Hongyoon Choi. Automated tumor microenvironment analysis for multiple samples by image-based spatial transcriptomics on tissue microarray [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4948.

Ex-Vivo Hippocampus Segmentation Using Diffusion-Weighted MRI

The hippocampus is a crucial brain structure involved in memory formation, spatial navigation, emotional regulation, and learning. An accurate MRI image segmentation of the human hippocampus plays an important role in multiple neuro-imaging research and clinical practice, such as diagnosing neurological diseases and guiding surgical interventions. While most hippocampus segmentation studies focus on using T1-weighted or T2-weighted MRI scans, we explore the use of diffusion-weighted MRI (dMRI), which offers unique insights into the microstructural properties of the hippocampus. Particularly, we utilize various anisotropy measures derived from diffusion MRI (dMRI), including fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity, for a multi-contrast deep learning approach to hippocampus segmentation. To exploit the unique benefits offered by various contrasts in dMRI images for accurate hippocampus segmentation, we introduce an innovative multimodal deep learning architecture integrating cross-attention mechanisms. Our proposed framework comprises a multi-head encoder designed to transform each contrast of dMRI images into distinct latent spaces, generating separate image feature maps. Subsequently, we employ a gated cross-attention unit following the encoder, which facilitates the creation of attention maps between every pair of image contrasts. These attention maps serve to enrich the feature maps, thereby enhancing their effectiveness for the segmentation task. In the final stage, a decoder is employed to produce segmentation predictions utilizing the attention-enhanced feature maps. The experimental outcomes demonstrate the efficacy of our framework in hippocampus segmentation and highlight the benefits of using multi-contrast images over single-contrast images in diffusion MRI image segmentation.

Balloon technologies for pulmonary vein isolation-12-month outcome and comparison of the novel radiofrequency balloon with the cryoballoon in patients with paroxysmal atrial fibrillation.

The cryoballoon (CB) has become a standard tool for pulmonary vein isolation (PVI), but the technology is limited in certain ways. A novel RF-balloon (Heliostar™, Biosense Webster, CA, USA) promises the advantages of a balloon technology in combination with 3D mapping. To assess procedural data and outcome, all patients undergoing RF-balloon PVI were included and compared with data from consecutive patients undergoing CB PVI for paroxysmal AF. A total of 254 patients (63 ± 13years, 54% male) were included: 30 patients undergoing RF-balloon and 224 patients CB PVI. Baseline parameters were comparable. Procedure duration (104.3 ± 35.3min vs. 69.9 ± 23.1min; p ≤ 0.001) and fluoroscopy time (16.3 ± 7.1min vs. 11.6 ± 4.9min; p ≤ 0.001) were longer using the RF-balloon; ablation time (43.5 ± 17.9 vs. 36.4 ± 15.6; p = 0.08) did not differ, and time-to-isolation (TTI) was shorter (18.2 ± 7.0s vs. 62.8 ± 35.1s; p ≤ 0.001). Second-generation RF-balloon cases showed shorter ablation time and TTI at comparable procedure duration and fluoroscopy time. One pericardial effusion occurred with the RF-balloon due to complicated transseptal access. During CB PVI in 4/224 patients (1.8%), a phrenic nerve palsy was observed. After 12months, 78% of patients after RF-balloon and 81% of patients after CB PVI (p = 0.5) were free from atrial arrhythmias. The RF-balloon was safe and effective. Compared with the CB, TTI was shorter, but procedure durations and fluoroscopy times were longer. This can be attributed to a learning curve and the initial necessity for separate 3D map preparation. Considering the results with the second-generation RF-balloon, more experience is needed to determine the potential benefits.