Memory Constraints Research Articles

Comparison of 2D, 2.5D, and 3D segmentation networks for maxillary sinuses and lesions in CBCT images

BackgroundThe purpose of this study was to compare the segmentation performances of the 2D, 2.5D, and 3D networks for maxillary sinuses (MSs) and lesions inside the maxillary sinus (MSL) with variations in sizes, shapes, and locations in cone beam CT (CBCT) images under the same constraint of memory capacity.MethodsThe 2D, 2.5D, and 3D networks were compared comprehensively for the segmentation of the MS and MSL in CBCT images under the same constraint of memory capacity. MSLs were obtained by subtracting the prediction of the air region of the maxillary sinus (MSA) from that of the MS.ResultsThe 2.5D network showed the highest segmentation performances for the MS and MSA compared to the 2D and 3D networks. The performances of the Jaccard coefficient, Dice similarity coefficient, precision, and recall by the 2.5D network of U-net + + reached 0.947, 0.973, 0.974, and 0.971 for the MS, respectively, and 0.787, 0.875, 0.897, and 0.858 for the MSL, respectively.ConclusionsThe 2.5D segmentation network demonstrated superior segmentation performance for various MSLs with an ensemble learning approach of combining the predictions from three orthogonal planes.

Machine learning framework for Inter-Beat Interval estimation using wearable Photoplethysmography sensors

Accurate estimation of Heart Rate (HR) is crucial for various health and fitness tracking applications in wearable devices. Among the existing techniques, those based on Photoplethysmography (PPG) signals have been extensively explored for continuous HR monitoring. In this paper, we introduce a new framework that utilizes machine learning methods to estimate Inter-Beat Interval (IBI) from wrist PPG signals. The primary objective of the proposed methods is to enhance the accuracy of IBI estimates while ensuring their feasibility for wearable devices, considering typical computational power, and memory constraints. This is achieved through two combined approaches. The first method involves optimizing a set of parameters associated with the interpolation of the systolic peaks. During the training stage of this method, we optimize the weights applied to the neighboring samples of the systolic peaks while minimizing the error between the fine-tuned peak positions and their corresponding ground truth. The second method is related to a supervised machine learning approach that handles the R-peak detection as a classification task. This allows us to accurately detect systolic peaks in PPG signals by classifying candidate peaks as true or false systolic peaks. To evaluate the effectiveness of our proposed methods, we conducted experiments on a dataset collected and carefully organized to this end. It consists of PPG and Electrocardiogram (ECG) signals obtained from 46 volunteers, including individuals with normal sinus rhythm, atrial fibrillation, and other non-specified cardiac arrhythmias. Comparing our framework with other state-of-the-art methods, we observed excellent performance, with lower Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) in IBI estimation. The proposed approach also achieved higher levels of Precision and Recall in terms of the detection of systolic peaks.

Periodic fluctuations in reading times reflect multi-word-chunking

Memory is fleeting. To avoid information loss, humans need to recode verbal stimuli into chunks of limited duration, each containing multiple words. Chunk duration may also be limited neurally by the wavelength of periodic brain activity, so-called neural oscillations. While both cognitive and neural constraints predict some degree of behavioral regularity in processing, this remains to be shown. Our analysis of self-paced reading data from 181 participants reveals periodic patterns at a frequency of sim 2 Hz. We defined multi-word chunks by using a computational formalization based on dependency annotations and part-of-speech tags. Potential chunk outputs were first generated from the computational formalization and the final chunk outputs were selected based on normalized pointwise mutual information. We show that behavioral periodicity is time-aligned to multi-word chunks, suggesting that the multi-word chunks generated from local dependency clusters may minimize memory demands. This is the first evidence that sentence processing behavior is periodic, consistent with a role of both memory constraints and endogenous electrophysiological rhythms in the formation of chunks during language comprehension.

Fast Parabolic Fitting: An R-Peak Detection Algorithm for Wearable ECG Devices.

Heart diseases rank among the most fatal health concerns globally, with the majority being preventable through early diagnosis and effective treatment. Electrocardiogram (ECG) analysis is critical in detecting heart diseases, as it captures the heart's electrical activities. For continuous monitoring, wearable electrocardiographic devices must ensure user comfort over extended periods, typically 24 to 48 h. These devices demand specialized algorithms with low computational complexity to accommodate memory and power consumption constraints. One of the most crucial aspects of ECG signals is accurately detecting heartbeat intervals, specifically the R peaks. In this study, we introduce a novel algorithm designed for wearable devices, offering two primary attributes: robustness against noise and low computational complexity. Our algorithm entails fitting a least-squares parabola to the ECG signal and adaptively shaping it as it sweeps through the signal. Notably, our proposed algorithm eliminates the need for band-pass filters, which can inadvertently smooth the R peaks, making them more challenging to identify. We compared the algorithm's performance using two extensive databases: the meta-database QT database and the BIH-MIT database. Importantly, our method does not necessitate the precise localization of the ECG signal's isoelectric line, contributing to its low computational complexity. In the analysis of the QT database, our algorithm demonstrated a substantial advantage over the classical Pan-Tompkins algorithm and maintained competitiveness with state-of-the-art approaches. In the case of the BIH-MIT database, the performance results were more conservative; they continued to underscore the real-world utility of our algorithm in clinical contexts.

Developing an implementation intervention for managing acute vertigo in the emergency department

BackgroundThere are evidence-based bedside tests for diagnosing acute vertigo, but no evidence-based strategies to support clinicians in implementing them. The purpose of this study was to design an implementation strategy...

Mondrian forest for data stream classification under memory constraints

Mondrian forest for data stream classification under memory constraints

Specific quinone reductase 2 inhibitors reduce metabolic burden and reverse Alzheimer's disease phenotype in mice.

Biological aging can be described as accumulative, prolonged metabolic stress and is the major risk factor for cognitive decline and Alzheimer's disease (AD). Recently, we identified and described a quinone reductase 2 (QR2) pathway in the brain, in which QR2 acts as a removable memory constraint and metabolic buffer within neurons. QR2 becomes overexpressed with age, and it is possibly a novel contributing factor to age-related metabolic stress and cognitive deficit. We found that, in human cells, genetic removal of QR2 produced a shift in the proteome opposing that found in AD brains while simultaneously reducing oxidative stress. We therefore created highly specific QR2 inhibitors (QR2is) to enable evaluation of chronic QR2 inhibition as a means to reduce biological age-related metabolic stress and cognitive decline. QR2is replicated results obtained by genetic removal of QR2, while local QR2i microinjection improved hippocampal and cortical-dependent learning in rats and mice. Continuous consumption of QR2is in drinking water improved cognition and reduced pathology in the brains of AD-model mice (5xFAD), with a noticeable between-sex effect on treatment duration. These results demonstrate the importance of QR2 activity and pathway function in the healthy and neurodegenerative brain and what we believe to be the great therapeutic potential of QR2is as first-in-class drugs.

Adaptive Sampling and Quick Anomaly Detection in Large Networks

The monitoring of data streams with a network structure have drawn increasing attention due to its wide applications in modern process control. In these applications, high-dimensional sensor nodes are interconnected with an underlying network topology. In such a case, abnormalities occurring to any node may propagate dynamically across the network and cause changes of other nodes over time. Furthermore, high dimensionality of such data significantly increased the cost of resources for data transmission and computation, such that only partial observations can be transmitted or processed in practice. Overall, how to quickly detect abnormalities in such large networks with resource constraints remains a challenge, especially due to the sampling uncertainty under the dynamic anomaly occurrences and network-based patterns. In this paper, we incorporate network structure information into the monitoring and adaptive sampling methodologies for quick anomaly detection in large networks where only partial observations are available. We develop a general monitoring and adaptive sampling method and further extend it to the case with memory constraints, both of which exploit network distance and centrality information for better process monitoring and identification of abnormalities. Theoretical investigations of the proposed methods demonstrate their sampling efficiency on balancing between exploration and exploitation, as well as the detection performance guarantee. Numerical simulations and a case study on power network have demonstrated the superiority of the proposed methods in detecting various types of shifts. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —Continuous monitoring of networks for anomalous events is critical for a large number of applications involving power networks, computer networks, epidemiological surveillance, social networks, etc. This paper aims at addressing the challenges in monitoring large networks in cases where monitoring resources are limited such that only a subset of nodes in the network is observable. Specifically, we integrate network structure information of nodes for constructing sequential detection methods via effective data augmentation, and for designing adaptive sampling algorithms to observe suspicious nodes that are likely to be abnormal. Then, the method is further generalized to the case that the memory of the computation is also constrained due to the network size. The developed method is greatly beneficial and effective for various anomaly patterns, especially when the initial anomaly randomly occurs to nodes in the network. The proposed methods are demonstrated to be capable of quickly detecting changes in the network and dynamically changes the sampling priority based on online observations in various cases, as shown in the theoretical investigation, simulations and case studies.

SketchNE: Embedding Billion-Scale Networks Accurately in One Hour

We study large-scale network embedding with the goal of generating high-quality embeddings for networks with more than 1 billion vertices and 100 billion edges. Recent attempts LightNE and NetSMF propose to sparsify and factorize the (dense) NetMF matrix for embedding large networks, where NetMF is a theoretically-grounded network embedding method. However, there is a trade-off between their embeddings' quality and scalability due to their expensive memory requirements, making embeddings less effective under real-world memory constraints. Therefore, we present the SketchNE model, a scalable, effective, and memory-efficient network embedding solution developed for a single machine with CPU only. The main idea of SketchNE is to avoid the explicit construction and factorization of the NetMF matrix either sparsely or densely when producing the embeddings through the proposed sparse-sign randomized single-pass SVD algorithm. We conduct extensive experiments on nine datasets of various sizes for vertex classification and link prediction, demonstrating the consistent outperformance of SketchNE over state-of-the-art baselines in terms of both effectiveness and efficiency. SketchNE costs only <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1.0 hours</b> to embed the Hyperlink2012 network with <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3.5 billion</b> vertices and <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">225 billion</b> edges on a CPU-only single machine with embedding superiority (e.g., a <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">282%</b> relative HITS@10 gain over LightNE).

FireFace: Leveraging Internal Function Features for Configuration of Functions on Serverless Edge Platforms.

The emerging serverless computing has become a captivating paradigm for deploying cloud applications, alleviating developers' concerns about infrastructure resource management by configuring necessary parameters such as latency and memory constraints. Existing resource configuration solutions for cloud-based serverless applications can be broadly classified into modeling based on historical data or a combination of sparse measurements and interpolation/modeling. In pursuit of service response and conserving network bandwidth, platforms have progressively expanded from the traditional cloud to the edge. Compared to cloud platforms, serverless edge platforms often lead to more running overhead due to their limited resources, resulting in undesirable financial costs for developers when using the existing solutions. Meanwhile, it is extremely challenging to handle the heterogeneity of edge platforms, characterized by distinct pricing owing to their varying resource preferences. To tackle these challenges, we propose an adaptive and efficient approach called FireFace, consisting of prediction and decision modules. The prediction module extracts the internal features of all functions within the serverless application and uses this information to predict the execution time of the functions under specific configuration schemes. Based on the prediction module, the decision module analyzes the environment information and uses the Adaptive Particle Swarm Optimization algorithm and Genetic Algorithm Operator (APSO-GA) algorithm to select the most suitable configuration plan for each function, including CPU, memory, and edge platforms. In this way, it is possible to effectively minimize the financial overhead while fulfilling the Service Level Objectives (SLOs). Extensive experimental results show that our prediction model obtains optimal results under all three metrics, and the prediction error rate for real-world serverless applications is in the range of 4.25∼9.51%. Our approach can find the optimal resource configuration scheme for each application, which saves 7.2∼44.8% on average compared to other classic algorithms. Moreover, FireFace exhibits rapid adaptability, efficiently adjusting resource allocation schemes in response to dynamic environments.

Online educational video engagement prediction based on dynamic graph neural networks

PurposeWith the continuous growth of online education, the quality issue of online educational videos has become increasingly prominent, causing students in online learning to face the dilemma of knowledge confusion. The existing mechanisms for controlling the quality of online educational videos suffer from subjectivity and low timeliness. Monitoring the quality of online educational videos involves analyzing metadata features and log data, which is an important aspect. With the development of artificial intelligence technology, deep learning techniques with strong predictive capabilities can provide new methods for predicting the quality of online educational videos, effectively overcoming the shortcomings of existing methods. The purpose of this study is to find a deep neural network that can model the dynamic and static features of the video itself, as well as the relationships between videos, to achieve dynamic monitoring of the quality of online educational videos.Design/methodology/approachThe quality of a video cannot be directly measured. According to previous research, the authors use engagement to represent the level of video quality. Engagement is the normalized participation time, which represents the degree to which learners tend to participate in the video. Based on existing public data sets, this study designs an online educational video engagement prediction model based on dynamic graph neural networks (DGNNs). The model is trained based on the video’s static features and dynamic features generated after its release by constructing dynamic graph data. The model includes a spatiotemporal feature extraction layer composed of DGNNs, which can effectively extract the time and space features contained in the video's dynamic graph data. The trained model is used to predict the engagement level of learners with the video on day T after its release, thereby achieving dynamic monitoring of video quality.FindingsModels with spatiotemporal feature extraction layers consisting of four types of DGNNs can accurately predict the engagement level of online educational videos. Of these, the model using the temporal graph convolutional neural network has the smallest prediction error. In dynamic graph construction, using cosine similarity and Euclidean distance functions with reasonable threshold settings can construct a structurally appropriate dynamic graph. In the training of this model, the amount of historical time series data used will affect the model’s predictive performance. The more historical time series data used, the smaller the prediction error of the trained model.Research limitations/implicationsA limitation of this study is that not all video data in the data set was used to construct the dynamic graph due to memory constraints. In addition, the DGNNs used in the spatiotemporal feature extraction layer are relatively conventional.Originality/valueIn this study, the authors propose an online educational video engagement prediction model based on DGNNs, which can achieve the dynamic monitoring of video quality. The model can be applied as part of a video quality monitoring mechanism for various online educational resource platforms.

A Sequential Addressing Subsampling Method for Massive Data Analysis Under Memory Constraint

The emergence of massive data in recent years brings challenges to automatic statistical inference. This is particularly true if the data are too numerous to be read into memory as a whole. Accordingly, new sampling techniques are needed to sample data from a hard drive. In this paper, we propose a sequential addressing subsampling (SAS) method that can sample data directly from the hard drive. The newly proposed SAS method is time saving in terms of addressing cost compared to that of the random addressing subsampling (RAS) method. Estimators (e.g., the sample mean) based on the SAS subsamples are constructed, and their properties are studied. We conduct a series of simulation studies to verify the finite sample performance of the proposed SAS estimators. The time cost is also compared between the SAS and RAS methods. An analysis of the airline data is presented for illustration purpose.

AutoMap: Automatic Mapping of Neural Networks to Deep Learning Accelerators for Edge Devices

Emerging deep neural networks (DNNs) have been emerging in applications (object detection, automatic speech recognition, etc.) deployed on edge devices. To improve the energy efficiency of edge devices, domain specific deep learning accelerators (DLAs) are designed with limited on-chip resources. The manifold DLA designs and evolving DNN topologies bring challenges for applications mapping and scheduling on hardware resources. In this paper, we propose an automatic DNN mapping framework named AutoMap given the hardware backend information. Firstly, a computational graph representation called Extended Directed Weighted Graph (EDWG) is proposed, which realizes unified expression for both spatial and temporal network inter-layer connections. Secondly, an associated partitioner is implemented for splitting an EDWG into subEDWGs, which incorporates the on-chip memory constraint and facilitates weight data reuse on chip. Lastly, a dynamic memory allocation strategy is utilized to alleviate the feature storing burden introduced by the multivarious network sizes and connections. Compared to the baseline mapping methods, experimental results show that our proposed automatic mapping framework can help to speedup the execution of several DNNs on state-of-the-art DLAs, ranging from 1.27x to 3.45x. The utilization of the PE array can increase from 20% to 64%.

Efficient construction of the BWT for repetitive text using string compression

We present a new semi-external algorithm that builds the Burrows–Wheeler transform variant of Bauer et al. (a.k.a., BCR BWT) in linear expected time. Our method uses compression techniques to reduce computational costs when the input is massive and repetitive. Concretely, we build on induced suffix sorting (ISS) and resort to run-length and grammar compression to maintain our intermediate results in compact form. Our compression format not only saves space but also speeds up the required computations. Our experiments show important space and computation time savings when the text is repetitive. In moderate-size collections of real human genome assemblies (14.2 GB - 75.05 GB), our memory peak is, on average, 1.7x smaller than the peak of the state-of-the-art BCR BWT construction algorithm (ropebwt2), while running 5x faster. Our current implementation was also able to compute the BCR BWT of 400 real human genome assemblies (1.2 TB) in 41.21 hours using 118.83 GB of working memory (around 10% of the input size). Interestingly, the results we report in the 1.2 TB file are dominated by the difficulties of scanning huge files under memory constraints (specifically, I/O operations). This fact indicates we can perform much better with a more careful implementation of our method, thus scaling to even bigger sizes efficiently.

Advancing Robot Perception in Non-Spiral Environments through Camera-based Image Processing

Robot perception heavily relies on camera-based visual input for navigating and interacting in its environment. As robots become integral parts of various applications, the need to efficiently compute their visibility regions in complex environments has grown. The key challenge addressed in this paper is to devise an innovative solution that not only accurately computes the visibility region V of a robot operating in a polynomial environment but also optimizes memory utilization to ensure real-time performance and scalability.The main objective of this research is to propose an algorithm that achieves optimal-time complexity and significantly reduces memory requirements for visibility region computation. By focusing on sub-linear memory utilization, we aim to enhance the robot's ability to perceive its surroundings effectively and efficiently.Previous approaches have provided solutions for visibility region computation in non-spiral environments, but most were not tailored to memory limitations. In contrast, the proposed algorithm is designed to achieve optimal time complexity that is O(n) while reducing memory usage to O(c/log n) variables, where c &lt; n represents the number of critical corners in the environment. Leveraging the constant-memory model and memory-constrained algorithm, we aim to strike a balance between computational efficiency and memory usage.The algorithm's performance is rigorously evaluated through extensive simulations and practical experiments. The results demonstrate its linear-time complexity and substantial reduction in memory usage without compromising the accuracy of the visibility region computation. By efficiently handling memory constraints, the robot gains a cost-effective and reliable perception mechanism, making it well-suited for a wide range of real-world applications.The constant-memory model and memory-constrained algorithm presented in this paper offer a significant advancement in robot perception capabilities. By optimizing the visibility region computation in polynomial environments, our approach contributes to the efficient operation of robots, enhancing their performance and applicability in complex real-world scenarios. The results of this research hold promising potential for future developments in robotics, computer vision, and related fields.

Advanced road extraction using CNN-based U-Net model and satellite imagery

Road extraction from high-resolution remote sensing imagery (HRRSI) is a challenging task due to low spectral variation and the presence of complex background elements, such as the shadows of buildings and trees. Many techniques have struggled to maintain proper edges and boundaries while retaining the geometric features necessary for accurate non-linear road extraction. In this paper, we address these issues by proposing a deep learning approach that utilizes a fine-tuned U-Net model with a modified feature space in the basic U-Net architecture for semantic segmentation. We also incorporate the BRISQUE preprocessing technique to improve the performance. We experimented with a subset of 200 high-quality images from the Massachusetts roads dataset, prioritized based on their rank. Due to computer memory constraints, the images were resized to 256 by 256 pixels. The proposed method produced the accuracy of 95.45%.

Concept drift detection and adaption framework using optimized deep learning and adaptive sliding window approach

AbstractConcept drift in online streaming data is a common issue due to dynamic smart systems, which results in system failure or performance degradation. Though there are several traditional approaches for handling the streaming data, they failed to handle the concept drift imposing the need for developing an adaptable approach to managing the dynamic IoT streaming data. Therefore, in this research, a new method is proposed for handling the concept drift issues in online data streaming. This research develops the dynamic streaming data analytic framework based on the optimized Deep CNN and optimized adaptive and sliding window (OASW) approach that effectively addresses both memory and time constraints. An optimized Deep CNN classifier is employed as a base classifier for offline learning, which is developed through hybridizing the proposed Desale's aggressive hunt optimization (AHO) algorithm with a Deep CNN classifier for tuning the optimal parameters of the classifier. An optimized adaptive and sliding window is utilized in this research to adapt the pattern changes in the data streams, which effectively handles the concept drift. The experimental analysis reveals that the proposed methods outperform the conventional methods considered for the analysis in terms of specificity, sensitivity, accuracy, F1 score, and the precision score of 96.65%, 97.77%, 98.63%, 98.1487%, and 98.4469%, respectively.

Searching with Distributional Heuristics (Student Abstract)

Distributional heuristic search uses distributions rather than point values for its heuristic estimates of the cost to goal from any given state. Distributional heuristics are desirable as they provide search algorithms not only with a way to evaluate nodes, but also with a basis for rational decision making tailored to specific search settings. Bounded suboptimal, anytime, and contract searches have differing but related objectives that each lend themselves to probabilistic reasoning supported by distributional heuristics. In many applications, speed of planning can be more important than solution quality. Whether due to certain domains' inherent difficulty, where anything but a satisficing approach is infeasible due to time or memory constraints, or due to the limited planning time available in real-time robotics and other time-sensitive planning settings, important open questions are how best to find solutions as quickly as possible and how to find the best solution possible while subject to an explicit limit on planning time. Successful algorithms must reason not only about solution cost, possibly in relation to a suboptimality bound, but also about the relative likelihood of finding a solution under one node vs. under another, of finding a solution of a particular cost (such as in relation to that of an incumbent solution), or about the expected amount of search effort to find a goal under a given node. This dissertation takes up these issues in four parts. I (1) examine different methods for generating distributional heuristics in bounded cost heuristic search and classical planning; (2) study the contract search setting, which involves online estimation of several unknown values; (3) consider the bounded suboptimal setting; and (4) address the anytime setting.

Optimizing the operation of photovoltaic panels using mathematical regression models

Solar (PV) panel models are necessary for the implementation of solar panel control algorithms such as maximum power point tracking. If real-time control is required, these models will be used on microcontrollers and, therefore, will be subject to constraints of memory and execution time: the model implemented on the microcontroller must occupy a memory area as small as possible and the mathematical relationships it contains can be computed in the shortest possible time. Analytical models usually do not fit this criteria, therefore searching for equivalent models which do fit it is required. In this paper, equivalent mathematical models are determined in the MATLAB programming environment based on data obtained by simulating an analytical model in the LabVIEW programming environment. In this paper, the models were obtained by modeling the current distribution in the panel through regression methods, more specifically by using an optimal response surface which has been validated through the resolution of an optimization problem with restrictions. The optimal response surface was used to estimate the extreme values of the power produced by the panel for certain given weather conditions (maximum power in summer and minimum power in winter).

An Improved Combination of Faster R-CNN and U-Net Network for Accurate Multi-Modality Whole Heart Segmentation.

Detailed information of substructures of the whole heart is usually vital in the diagnosis of cardiovascular diseases and in 3D modeling of the heart. Deep convolutional neural networks have been demonstrated to achieve state-ofthe-art performance in 3D cardiac structures segmentation. However, when dealing with high-resolution 3D data, current methods employing tiling strategies usually degrade segmentation performances due to GPU memory constraints. This work develops a two-stage multi-modality whole heart segmentation strategy, which adopts an improved Combination of Faster R-CNN and 3D U-Net (CFUN+). More specifically, the bounding box of the heart is first detected by Faster R-CNN, and then the original Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) images of the heart aligned with the bounding box are input into 3D U-Net for segmentation. The proposed CFUN+ method redefines the bounding box loss function by replacing the previous Intersection over Union (IoU) loss with Complete Intersection over Union (CIoU) loss. Meanwhile, the integration of the edge loss makes the segmentation results more accurate, and also improves the convergence speed. The proposed method achieves an average Dice score of 91.1% on the Multi-Modality Whole Heart Segmentation (MM-WHS) 2017 challenge CT dataset, which is 5.2% higher than the baseline CFUN model, and achieves state-of-the-art segmentation results. In addition, the segmentation speed of a single heart has been dramatically improved from a few minutes to less than 6 seconds.