Down-sampled Data Research Articles

Effect of temporal resolution on the detection of cardiac arrhythmias using HRV features and machine learning

Objective. Arrhythmia is an abnormal cardiac rhythm that affects the pattern and rate of the heartbeat. Wearable devices with the functionality to measure and store heart rate (HR) data are growing in popularity and enable diagnosing and monitoring arrhythmia on a large scale. The typical sampling resolution of HR data available from non-medical grade wearable devices varies from seconds to several minutes depending on the device and its settings. However, the impact of sampling resolution on the performance and quality of arrhythmia detection has not yet been quantified. Approach. In this study, we investigated the detection and classification of three arrhythmias, namely atrial fibrillation, bradycardia, tachycardia, from down-sampled HR data with various temporal resolution (5-, 15-, 30- and 60 s averages) in 1 h segments extracted from an annotated Holter ECG database acquired at the University of Virginia Heart Station. For the classification task, a total of 15 common heart rate variability (HRV) features were engineered based on the HR time series of each patient. Three different types of machine learning classifiers were evaluated, namely logistic regression, support vector machine and random forest. Main results. A decrease in temporal resolution drastically impacted the detection of atrial fibrillation but did not substantially affect the detection of bradycardia and tachycardia. A HR resolution up to 15 s average demonstrated reasonable performance with a sensitivity of 0.92 and a specificity of 0.86 for a multiclass random forest classifier. Significance. HRV features extracted from low resolution long HR recordings have the potential to increase the early detection of arrhythmias in undiagnosed individuals.

Frequency-band down-sampled stereo-DIC: Beyond the limitation of single frequency excitation

A low-speed stereo-camera DIC setup was used in this paper to measure the down-sampled bandpass vibration signal of a plate in a given frequency range, which is much higher than the available frame rate. Down-sampled vibration measurements are well known in the literature. However, they are always subject to the strong hypothesis of having a single frequency component in the excitation source (e.g., pure sinusoidal excitation). In this scenario, several approaches can be found in the literature to reconstruct the actual response from the down-sampled data. In this paper, a data postprocessing algorithm is newly introduced to properly reconstruct the target response in the case of a frequency band excitation, thus relaxing the single frequency excitation constraint and allowing to explore a given frequency range with a single measurement. Additionally, a custom excitation signal is presented in this paper to achieve a constant intensity in the studied frequency range. The proposed approach is presented and experimentally validated by measuring a cantilever plate vibration in the kHz range using a 100 fps acquisition. A single acquisition of 98 frames allowed to describe the deformed shapes at 49 different frequency values in the range 1110–1160 Hz, highlighting two resonance peaks. The comparison with the results of multiple conventional single-frequency tests and LDV measurements confirmed the effectiveness of the proposed approach.

Lithography Hotspot Detection Method Based on Transfer Learning Using Pre-Trained Deep Convolutional Neural Network

As the designed feature size of integrated circuits (ICs) continues to shrink, the lithographic printability of the design has become one of the important issues in IC design and manufacturing. There are patterns that cause lithography hotspots in the IC layout. Hotspot detection affects the turn-around time and the yield of IC manufacturing. The precision and F1 score of available machine-learning-based hotspot-detection methods are still insufficient. In this paper, a lithography hotspot detection method based on transfer learning using pre-trained deep convolutional neural network is proposed. The proposed method uses the VGG13 network trained with the ImageNet dataset as the pre-trained model. In order to obtain a model suitable for hotspot detection, the pre-trained model is trained with some down-sampled layout pattern data and takes cross entropy as the loss function. ICCAD 2012 benchmark suite is used for model training and model verification. The proposed method performs well in accuracy, recall, precision, and F1 score. There is significant improvement in the precision and F1 score. The results show that updating the weights of partial convolutional layers has little effect on the results of this method.

Deep neural networks for magnetic resonance elastography acceleration in thermal-ablation monitoring.

To develop a deep neural network for accelerating magnetic resonance elastography (MRE) acquisition, to validate the ability to generate reliable MRE results with the down-sampled k-space data, and to demonstrate the feasibility of the proposed method in monitoring the stiffness changes during thermal ablation in a phantom study. MRE scans were performed with 60Hz excitation on porcine ex-vivo liver gel phantoms in a 0.36T MRI scanner to generate the training dataset. The acquisition protocol was based on a spin-echo MRE pulse sequence with tailored motion-sensitive gradients to reduce echo time (TE). A U-Net based deep neural network was developed and trained to interpolate the missing data from down-sampled k-space. We calculated the errors of 80 sets magnitude/phase images reconstructed from the zero-filled, compressive sensing (CS) and deep learning (DL) method for comparison. The peak signal-to-noise rate (PSNR) and structural similarity index (SSIM) of the magnitude/phase images were also calculated for comparison. The stiffness changes were recorded before, during, and after ablation. The mean stiffness values over the region of interest (ROI) were compared between the elastograms reconstructed from the fully sampled k-space and interpolated k-space after thermal ablation. The mean absolute error (MAE), PSNR, and SSIM of the proposed DL approach were significantly better than the results from the zero-filled method (p<0.0001) and CS (p<0.0001). The stiffness changes before and after thermal ablation assessed by the proposed approach (before: 7.7±1.1kPa, after: 11.9±4.0kPa, 4.2-kPa increase) gave close agreement with the values calculated from the fully sampled data (before: 8.0±1.0kPa, after: 12.6±4.2kPa, 4.6-kPa increase). In contrast, the stiffness changes computed from the zero-filled method (before: 4.9±1.4kPa, after: 5.6±2.8kPa, 0.7-kPa increase) were substantially underestimated. This study demonstrated the capability of the proposed DL method for rapid MRE acquisition and provided a promising solution for monitoring the MRI-guided thermal ablation.

An adaptive down-sampling method of laser scan data for scan-to-BIM

Laser scan data are popularly adopted to capture the conditions of existing buildings for as-is BIM reconstruction, also known as scan-to-BIM. Down-sampling of massive laser scan data is a critical pre-processing step for efficient scan-to-BIM. It is also crucial to accurately retain the geometric and semantic information of existing buildings during the down-sampling process. However, existing down-sampling approaches mainly focus on the preservation of geometric features only, without considering semantic features. This study developed an adaptive down-sampling method, which is able to maintain scan points containing critical geometric or semantic information and only down-sample scan points containing no critical information. The proposed method first conducts a geometry-based segmentation to identify edge points and non-planar points, which contain critical geometric information. Then, a semantic-based segmentation is performed to identify points with critical semantic information (e.g. labels and information boards). All the points containing either geometric or semantic information are retained, while points containing no critical information are down-sampled to reduce data redundancy. Experiments were conducted on four scenes, which showed that the proposed method was suitable for point cloud data with different accuracies and could achieve a better performance in preserving both geometric and semantic information than the traditional voxel-based method and the curvature-based method. It is proved that down-sampled data with the proposed method could generate as-is BIMs with more accurate geometric and semantic information.

3D multi-scale residual fully convolutional neural network for segmentation of extremely large-sized kidney tumor

Background and objectiveWe propose a novel deep neural network, the 3D Multi-Scale Residual Fully Convolutional Neural Network (3D-MS-RFCNN) to improve segmentation in extremely large-sized kidney tumors. MethodThe multi-scale approach with a deep neural network is applied to capture global contextual features. Our method, 3D-MS-RFCNN, consists of two encoders and one decoder as a single complete network. One of the encoders is designed for capturing global contextual information by using the low-resolution, down-sampled data from input images. In the decoder, features from the encoder for global contextual features are concatenated with up-sampled features from the previous layer and features from the other encoder. Ensemble learning strategy is also applied. ResultsWe evaluated the performance of our proposed method using the KiTS public dataset and the in-house hospital dataset. When compared with the state-of-the-art method, Res3D U-Net, our model, 3D-MS-RFCNN, demonstrated greater accuracy (0.9390 dice score for KiTS dataset and 0.8575 dice score for external dataset) for segmenting extremely large-sized kidney tumors. ConclusionsOur proposed network shows significantly improved segmentation performance of extremely large-sized targets. This study can be usefully employed in the field of medical image analysis.

Sparse Reconstruction for Radar Imaging Based on Quantum Algorithms

The sparse-driven radar imaging can obtain the high-resolution images about target scene with the down-sampled data. However, the huge computational complexity of the classical sparse recovery method for the particular situation seriously affects the practicality of the sparse imaging technology. In this paper, this is the first time the quantum algorithms are applied to the image recovery for the radar sparse imaging. Firstly, the radar sparse imaging problem is analyzed and the calculation problem to be solved by quantum algorithms is determined. Then, the corresponding quantum circuit and its parameters are designed to ensure extremely low computational complexity, and the quantum-enhanced reconstruction algorithm for sparse imaging is proposed. Finally, the computational complexity of the proposed method is analyzed, and the simulation experiments with the raw radar data are illustrated to verify the validity of the proposed method.

Design and Performance Analysis of Sparse TOPS SAR Mode

Terrain observation by progressive scans (TOPS) is a novel and promising wide-swath synthetic aperture radar (SAR) imaging mode, which overcomes the scalloping effect existing in ScanSAR. Traditional imaging algorithms are computationally efficient in raw data processing of TOPS, but the problems from clutter and sidelobes suppression have not been effectively solved. Sparse SAR imaging not only effectively solves these problems, but also has great potential in improving system performance, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">e.g.</i> , lower requirement of pulse repetition frequency (PRF), higher swath coverage, less data amount, and lower system complexity. Therefore, this paper proposes a novel sparse TOPS SAR imaging mode. The proposed mode combines advantages of sparse imaging system and TOPS mode to achieve high-quality and wide-swath SAR imaging. In this mode, we first design the beam position by timing diagram. Then the basic parameters of the mode are calculated according to the radar equation, and the properties such as grating lobes are analyzed. Finally, an <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{2,1}$</tex-math></inline-formula> -norm regularization based sparse TOPS SAR imaging algorithm is introduced to achieve the high-quality sparse scene recovery. Experimental results show that compared with conventional TOPS imaging mode, the proposed mode can obtain wide swath with lower PRF and effectively solve the effect on distributed target ambiguity ratio (DTAR). Compared with matched filtering (MF) based TOPS imaging algorithm, the sparse method can achieve unambiguous reconstruction of the considered scene from down-sampled data, and obtain the image with less artifacts.

Frequency-Scaling-Based Spaceborne Squint SAR Sparse Imaging

In synthetic aperture radar (SAR), due to the characteristics of squint data in spaceborne imaging geometry, a more accurate range cell migration correction (RCMC) operation is necessary. Furthermore, imaging degradation due to ignoring range-variant filtering needed for secondary range compression (SRC) limits SAR imaging. In addition, due to enormous computational complexity required for squint SAR data processing, the existing imaging methods, for instance, chirp scaling algorithm (CSA) and range-Doppler algorithm (RDA) are no longer sufficient, especially for large-scale scenes. In order to solve above problems, this paper presents a novel spaceborne squint SAR sparse imaging method, which could not only eliminate the effects of squint to a certain extent, but also improve the performance of focused SAR image. Compared with matched filtering (MF) based squint imaging algorithm, the proposed method can obtain the SAR images with higher quality from full- or down-sampled echo data.

Pupillary fluctuation amplitude before target presentation reflects short-term vigilance level in Psychomotor Vigilance Tasks.

Our daily activities require vigilance. Therefore, it is useful to externally monitor and predict our vigilance level using a straightforward method. It is known that the vigilance level is linked to pupillary fluctuations via Locus Coeruleus and Norepinephrine (LC-NE) system. However, previous methods of estimating long-term vigilance require monitoring pupillary fluctuations at rest over a long period. We developed a method of predicting the short-term vigilance level by monitoring pupillary fluctuation for a shorter period consisting of several seconds. The LC activity also fluctuates at a timescale of seconds. Therefore, we hypothesized that the short-term vigilance level could be estimated using pupillary fluctuations in a short period and quantified their amplitude as the Micro-Pupillary Unrest Index (M-PUI). We found an intra-individual trial-by-trial positive correlation between Reaction Time (RT) reflecting the short-term vigilance level and M-PUI in the period immediately before the target onset in a Psychomotor Vigilance Task (PVT). This relationship was most evident when the fluctuation was smoothed by a Hanning window of approximately 50 to 100 ms (including cases of down-sampled data at 100 and 50 Hz), and M-PUI was calculated in the period up to one or two seconds before the target onset. These results suggest that M-PUI can monitor and predict fluctuating levels of vigilance. M-PUI is also useful for examining pupillary fluctuations in a short period for elucidating the psychophysiological mechanisms of short-term vigilance.

On the down-sampling of diffusion MRI data along the angular dimension

BackgroundIt has been established that the diffusion gradient directions in diffusion MRI should be uniformly distributed in 3D spherical space, so that orientation-dependent diffusion properties (e.g., fractional anisotropy or FA) can be properly quantified. Sometimes the acquired data need to be down-sampled along the angular dimension before computing diffusion properties (e.g., to exclude data points corrupted by motion artifact; to harmonize data obtained with different protocols). It is important to quantitatively assess the impact of data down-sampling on measurement of diffusion properties. Materials and methodsHere we report 1) a numerical procedure for down-sampling diffusion MRI (e.g., for data harmonization), and 2) a spatial uniformity index of diffusion directions, aiming to predict the quality of the chosen down-sampling schemes (e.g., from data harmonization; or rejection of motion corrupted data points). We quantitatively evaluated human diffusion MRI data, which were down-sampled from 64 or 60 diffusion gradient directions to 30 directions, in terms of their 1) FA value accuracy (using fully-sampled data as the ground truth), 2) FA fitting residuals, and 3) spatial uniformity indices. ResultsOur experimental data show that the proposed spatial uniformity index is correlated with errors in FA obtained from down-sampled diffusion MRI data. The FA fitting residuals that are typically used to assess diffusion MRI quality are not correlated with either FA errors or spatial uniformity index. ConclusionsThese results suggest that the spatial uniformity index could be more valuable in assessing quality of down-sampled diffusion MRI data, as compared with FA fitting residual measures. We expect that our implemented software procedure should prove valuable for 1) guiding data harmonization for multi-site diffusion MRI studies, and 2) assessing the impact of rejecting motion corrupted data points on the accuracy of diffusion measures.

Sensitivity and specificity of CEST and NOE MRI in injured spinal cord in monkeys

PurposeThe sensitivity and accuracy of chemical exchange saturation transfer (CEST) and nuclear Overhauser enhancement (NOE) effects for assessing injury-associated changes in cervical spinal cords were evaluated in squirrel monkeys. Multiple interacting pools of protons, including one identified by an NOE at −1.6 ppm relative to water (NOE(-1.6)), were derived and quantified from fitting proton Z-spectra. The effects of down-sampled data acquisitions and corrections for non-specific factors including T1, semi-solid magnetization transfer, and direct saturation of free water (DS), were investigated. The overall goal is to develop a protocol for rapid data acquisition for assessing the molecular signatures of the injured spinal cord and its surrounding regions. MethodsMRI scans were recorded of anesthetized squirrel monkeys at 9.4 T, before and after a unilateral dorsal column sectioning of the cervical spinal cord. Z-spectral images at 51 different RF offsets were acquired. The amplitudes of CEST and NOE effects from multiple proton pools were quantified using a six-pool Lorenzian fitting of each Z-spectrum (MTRmfit). In addition, down-sampled data using reduced selections of RF offsets were analyzed and compared. An apparent exchange-dependent relaxation (AREXmfit) method was also used to correct for non-specific factors in quantifying regional spectra around lesion sites. ResultsThe parametric maps from multi-pool fitting using the complete sampling data (P51e) detected unilateral changes at and around the injury. The maps derived from selected twofold down-sampled data with appropriate interpolation (P26sI51) revealed quite similar spatial distributions of different pools as those obtained using P51e at each resonance shift. Across 10 subjects, both data acquisition schemes detected significant decreases in NOE(-3.5) and NOE(-1.6) and increases in DS(0.0) and CEST(3.5) at the lesion site relative to measures of the normal tissues before injury. AREXmfit of cysts and other abnormal tissues at and around the lesion site also exhibited significant changes, especially at 3.5, −1.6 and −3.5 ppm RF offsets. ConclusionThese results confirm that a reduced set of RF offsets and down sampling are adequate for CEST imaging of injured spinal cord and allow shorter imaging times and/or permit additional signal averaging. AREXmfit correction improved the accuracy of CEST and NOE measures. The results provide a rapid (~13 mins), sensitive, and accurate protocol for deriving multiple NOE and CEST effects simultaneously in spinal cord imaging at high field.

SPB-Net: A Deep Network for SAR Imaging and Despeckling With Downsampled Data

Synthetic aperture radar (SAR) typically faces both large-scale data and speckle noise problems, which, respectively, induce enormous strains on transmission and storage and interfere with the analysis and interpretation of SAR images. To tackle these, we present a real-time processing deep network, called SPB-Net, to implement imaging and speckle suppression simultaneously. First, a novel imaging-despeckling observation model with the nonlogarithmic additive speckle noise is established. Subsequently, guided by the statistical properties of noise and sparse and detail-preserved requirements in SAR imaging and despeckling, we formulate an <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$L_{2}$ </tex-math></inline-formula> along with two <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$L_{1}$ </tex-math></inline-formula> regularizations as the fidelity, sparse, and image detail-preserved constraints, respectively. Convolution layers are employed to improve the feature representation capability in the latter <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$L_{1}$ </tex-math></inline-formula> term as well. Based on this, we construct a corresponding convex optimization problem. Then, the complex-valued split Bregman method, focusing on the complex-variable convex problem, is unfolded into a parameter-learnable and architecture-fixed SPB-Net to solve the proposed problem effectively and efficiently. Experimental results with the downsampled Radarsat-1 raw data demonstrate the validity in imaging and speckle suppression and the real-time processing capability of the proposed SPB-Net.

Performance Comparison between Two Established Microgrid Planning MILP Methodologies Tested On 13 Microgrid Projects

Mixed Integer Linear Programming (MILP) optimization algorithms provide accurate and clear solutions for Microgrid and Distributed Energy Resources projects. Full-scale optimization approaches optimize all time-steps of data sets (e.g., 8760 time-step and higher resolutions), incurring extreme and unpredictable run-times, often prohibiting such approaches for effective Microgrid designs. To reduce run-times down-sampling approaches exist. Given that the literature evaluates the full-scale and down-sampling approaches only for limited numbers of case studies, there is a lack of a more comprehensive study involving multiple Microgrids. This paper closes this gap by comparing results and run-times of a full-scale 8760 h time-series MILP to a peak preserving day-type MILP for 13 real Microgrid projects. The day-type approach reduces the computational time between 85% and almost 100% (from 2 h computational time to less than 1 min). At the same time the day-type approach keeps the objective function (OF) differences below 1.5% for 77% of the Microgrids. The other cases show OF differences between 6% and 13%, which can be reduced to 1.5% or less by applying a two-stage hybrid approach that designs the Microgrid based on down-sampled data and then performs a full-scale dispatch algorithm. This two stage approach results in 20–99% run-time savings.

Wavelet Spectral Time-Frequency Training of Deep Convolutional Neural Networks for Accurate Identification of Micro-Scale Sharp Wave Biomarkers in the Post-Hypoxic-Ischemic EEG of Preterm Sheep.

Neonatal hypoxic-ischemic encephalopathy (HIE) evolves over different phases of time during recovery. Some neuroprotection treatments are only effective for specific, short windows of time during this evolution of injury. Clinically, we often do not know when an insult may have started, and thus which phase of injury the brain may be experiencing. To improve diagnosis, prognosis and treatment efficacy, we need to establish biomarkers which denote phases of injury. Our pre-clinical research, using preterm fetal sheep, show that micro-scale EEG patterns (e.g. spikes and sharp waves), superimposed on suppressed EEG background, primarily occur during the early recovery from an HI insult (0-6 h), and that numbers of events within the first 2 h are strongly predictive of neural survival. Thus, real-time automated algorithms that could reliably identify EEG patterns in this phase will help clinicians to determine the phases of injury, to help guide treatment options. We have previously developed successful automated machine learning approaches for accurate identification and quantification of HI micro-scale EEG patterns in preterm fetal sheep post-HI. This paper introduces, for the first time, a novel online fusion strategy that employs a high-level wavelet-Fourier (WF) spectral feature extraction method in conjunction with a deep convolutional neural network (CNN) classifier for accurate identification of micro-scale preterm fetal sheep post-HI sharp waves in 1024Hz EEG recordings, along with 256Hz down-sampled data. The classifier was trained and tested over 4120 EEG segments within the first 2 hours latent phase recordings. The WF-CNN classifier can robustly identify sharp waves with considerable high-performance of 99.86% in 1024Hz and 99.5% in 256Hz data. The method is an alternative deep-structure approach with competitive high-accuracy compared to our computationally-intensive WS-CNN sharp wave classifier.

Deep Convolutional Neural Networks for the Accurate Identification of High-Amplitude Stereotypic Epileptiform Seizures in the Post-Hypoxic-Ischemic EEG of Preterm Fetal Sheep.

Neonatal seizures after birth may contribute to brain injury after an hypoxic-ischemic (HI) event, impaired brain development and a later life risk for epilepsy. Despite neural immaturity, seizures can also occur in preterm infants. However, surprisingly little is known about their evolution after an HI insult or patterns of expression. An improved understanding of preterm seizures will help facilitate diagnosis and prognosis and the implementation of treatments. This requires improved detection of seizures, including electrographic seizures. We have established a stable preterm fetal sheep model of HI that results in different types of post-HI seizures. These including the expression of epileptiform transients during the latent phase (0-6 h) of cerebral energy recovery, and bursts of high amplitude stereotypic evolving seizures (HAS) during the secondary phase of cerebral energy failure (∼6-72 h). We have previously developed successful automated machine-learning strategies for accurate identification and quantification of the evolving micro-scale EEG patterns (e.g. gamma spikes and sharp waves), during the latent phase. The current paper introduces, for the first time, a real-time approach that employs a 15-layer deep convolutional neural network (CNN) classifier, directly fed with the raw EEG time-series, to identify HAS in the 1024Hz and 256Hz down-sampled data in our preterm fetuses post-HI. The classifier was trained and tested using EEG segments during ∼6 to 48 hours post-HI recordings. The classifier accurately identified HAS with 98.52% accuracy in the 1024Hz and 97.78% in the 256Hz data. Clinical relevance-Results highlight the promising ability of the proposed CNN classifier for accurate identification of HI related seizures in the neonatal preterm brain, if further applied to the current 256Hz clinical recordings, in real-world.

TuckerMPI

Our goal is compression of massive-scale grid-structured data, such as the multi-terabyte output of a high-fidelity computational simulation. For such data sets, we have developed a new software package called TuckerMPI, a parallel C++/MPI software package for compressing distributed data. The approach is based on treating the data as a tensor, i.e., a multidimensional array, and computing its truncated Tucker decomposition, a higher-order analogue to the truncated singular value decomposition of a matrix. The result is a low-rank approximation of the original tensor-structured data. Compression efficiency is achieved by detecting latent global structure within the data, which we contrast to most compression methods that are focused on local structure. In this work, we describe TuckerMPI, our implementation of the truncated Tucker decomposition, including details of the data distribution and in-memory layouts, the parallel and serial implementations of the key kernels, and analysis of the storage, communication, and computational costs. We test the software on 4.5 and 6.7 terabyte data sets distributed across 100 s of nodes (1,000 s of MPI processes), achieving compression ratios between 100 and 200,000×, which equates to 99--99.999% compression (depending on the desired accuracy) in substantially less time than it would take to even read the same dataset from a parallel file system. Moreover, we show that our method also allows for reconstruction of partial or down-sampled data on a single node, without a parallel computer so long as the reconstructed portion is small enough to fit on a single machine, e.g., in the instance of reconstructing/visualizing a single down-sampled time step or computing summary statistics. The code is available at https://gitlab.com/tensors/TuckerMPI.

A fair fight between molecular marker types in a seascape genetics setting.

From its inception, population genetics has been nearly as concerned with the genetic data type-to which analyses are brought to bear-as it is with the analysis methods themselves. The field has traversed allozymes, microsatellites, segregating sites in multilocus alignments and, currently, single nucleotide polymorphisms (SNPs) generated by high-throughput genomic sequencing methods, primarily whole genome sequencing and reduced representation library (RRL) sequencing. As each emerging data type has gained traction, it has been compared to existing methods, based on its relative ability to discern population structural complexity at increasing levels of resolution. However, this is usually done by comparing the gold standard in one data type to the gold standard in the new data type. These gold standards frequently differ in power and in sampling density, both across a genome and throughout a spatial range. In this issue of Molecular Ecology, D'Aloia et al. apply the high-throughput approach as fully as possible to microsatellites, nuclear loci and SNPs genotyped through an RRL method; this is coupled with a spatially dense sampling scheme. Completing a battery of population genetics analyses across data types (including a series of down-sampled data sets), the authors find that SNP data are slightly more sensitive to fine-scale genetic structure, and the results are more resilient to down-sampling than microsatellites and nonrepetitive nuclear loci. However, their results are far from an unqualified victory for RRL SNP data over all previous data types: the authors note that modest additions to the microsatellites and nuclear loci data sets may provide the necessary analytical power to delineate the fine-scale genetic structuring identified by SNPs. As always, as the field begins to fully embrace the newest thing, good science reminds us that traditional data types are far from useless, especially when combined with a well-designed sampling scheme.

Reconstruction of multicontrast MR images through deep learning.

Magnetic resonance (MR) imaging with a long scan time can lead to degraded images due to patient motion, patient discomfort, and increased costs. For these reasons, the role of rapid MR imaging is important. In this study, we propose the joint reconstruction of multicontrast brain MR images from down-sampled data to accelerate the data acquisition process using a novel deep-learning network. Twenty-one healthy volunteers (female/male=7/14, age=26±4yr, range 22-35yr) and 16 postoperative patients (female/male=7/9, age=49±9yr, range 37-62yr) were scanned on a 3T whole-body scanner for prospective and retrospective studies, respectively, using both T1-weighted spin-echo (SE) and T2-weighted fast spin-echo (FSE) sequences. We proposed a network which we term "X-net" to reconstruct both T1- and T2-weighted images from down-sampled images as well as a network termed "Y-net" which reconstructs T2-weighted images from highly down-sampled T2-weighted images and fully sampled T1-weighted images. Both X-net and Y-net are composed of two concatenated subnetworks. We investigate optimal sampling patterns, the optimal patch size for augmentation, and the optimal acceleration factors for network training. An additional Y-net combined with a generative adversarial network (GAN) was also implemented and tested to investigate the effects of the GAN on the Y-net performance. Single- and joint-reconstruction parallel-imaging and compressed-sensing algorithms along with a conventional U-net were also tested and compared with the proposed networks. For this comparison, the structural similarity (SSIM), normalized mean square error (NMSE), and Fréchet inception distance (FID) were calculated between the outputs of the networks and fully sampled images. The statistical significance of the performance was evaluated by assessing the interclass correlation and in paired t-tests. The outputs from the two concatenated subnetworks were closer to the fully sampled images compared to those from one subnetwork, with this result showing statistical significance. Uniform down-sampling led to a statically significant improvement in the image quality compared to random or central down-sampling patterns. In addition, the proposed networks provided higher SSIM and NMSE values than U-net, compressed-sensing, and parallel-imaging algorithms, all at statistically significant levels. The GAN-based Y-net showed a better FID and more realistic images compared to a non-GAN-based Y-net. The performance capabilities of the networks were similar between normal subjects and patients. The proposed X-net and Y-net effectively reconstructed full images from down-sampled images, outperforming the conventional parallel-imaging, compressed-sensing and U-net methods and providing more realistic images in combination with a GAN. The developed networks potentially enable us to accelerate multicontrast anatomical MR imaging in routine clinical studies including T1-and T2-weighted imaging.

FMCW SAR Sparse Imaging Based on Approximated Observation: An Overview on Current Technologies

Sparsity-driven synthetic aperture radar (SAR) imaging technique for frequency modulation continuous wave (FMCW) has already shown the superiority in terms of performance improvement in imaging and recovery from down-sampled data. However, restricted by the computational cost, conventional FMCW SAR sparse imaging method based on observation matrix is not able to achieve the large-scale scene reconstruction, not to mention real-time processing. To solve this problem, the FMCW SAR sparse imaging theory based on approximated observation is proposed by using an echo simulation operator to replace typical observation matrix, and recovering the scene via 2-D regularization operation. This new technology can achieve high-resolution sparse imaging of the scene with a computational cost close to that of traditional matched filtering algorithms, which makes several applications, such as early-warning and battlefield monitoring, possible by using FMCW SAR sparse imaging system. In this article, we present the recent research progress on approximated observation-based FMCW SAR sparse imaging to deal with a few key issues for practical radar systems. In particular, we describe: first, $L_q$ -norm $(0 regularization-based imaging technique that makes sparse reconstruction of large-scale scene possible; second, $L_{2,q}$ -norm regularization-based imaging technique that minimizes the azimuth ambiguities in high-resolution sparse imaging; and third, a sparse imaging technique that supports real-time applications of FMCW SAR imaging.