Compression Decompression Research Articles

A broad appraisal of decompression-induced physiological stress in different simulated dive profiles

Background: The present study was designed to observe if different decompression profiles, calculated as a function of tissue supersaturation during ascent, would result in significantly different outcomes, measured through different physiological stress indicators, even in the absence of symptoms of decompression sickness. Aim: The aim of this study was to evaluate if simulated decompression profiles would affect the immune system, oxidative stress indicators, and heart rate variability. Methods: A total of 23 volunteers participated in two different experimental protocols in a dry hyperbaric chamber. These simulated dives comprised two different compression–decompression arrangements with the same maximum pressure and duration but different decompression profiles. Results: The shallow decompression profile with shorter deeper stops and longer shallow stops presented an increase in the standard deviation of the normal-to-normal R-R interval (a wide indicator of overall variability); the deep decompression profile with longer deeper stops and shorter shallow stops did not exhibit such increase. The shallow decompression profile resulted in an increase in neutrophil count and its microparticles (MPs), but no changes were observed for platelet count and its MPs, as well as for endothelial-derived MPs. In contrast, the deep decompression profile resulted in no changes in neutrophil count and its MPs, but a decrease in platelet count along with an increase in MPs from both platelets and endothelial cells. The observed difference might be related to different levels of decompression-related activation of immune system responses and oxidative processes triggered by different levels of inert gas supersaturation upon surfacing. Conclusion: From previous results and literature data, we present a tentative schematic of how the velocity of ascent would trigger (or not) pro-inflammatory and immune system responses that could ultimately lead to the development of decompression sickness. Relevance for patients: Increasing safety in exposure to hyperbaric environments and subsequent decompression by evaluating individual physiological responses to the process.

Bio-mimicking DNA fingerprint profiling for HLS watermarking to counter hardware IP piracy

The multifaceted, multivendor-based global design supply chain induces hardware threats of intellectual property (IP) piracy for modern computing and electronic systems. Current hardware watermarking techniques fall short either in terms of watermark strength (size of covert constraints generated) or number of security layers/variables involved in the security constraints generation process. This paper presents a novel approach for high level synthesis (HLS) watermarking by bio-mimicking DNA fingerprint profiling to counter hardware IP piracy. The proposed approach effectively captures the vital DNA fingerprint profiling phases such as DNA sequencing, DNA fragmentation, fragment replication, DNA ligase, etc. and bio-mimics them to generate a digital watermarking framework. The presented approach has been demonstrated on convolutional layer and JPEG compression-decompression (CODEC) algorithms that are widely used in several medical and machine learning applications. The proposed approach has been thoroughly compared with several state-of-the-art approaches. The proposed approach depicts superior security in the probability of coincidence of up to ~ 104 and tamper tolerance of up to ~ 10368 at 0% overhead as compared to the prior approaches.

DermCompressNet: integrated CD-ConvNet and discrete cosine transform for dermoscopic images compression

AbstractTelemedicine has a critical role in healthcare by supporting the information exchange between the patients and the physicians as well as between the physicians for consultation. Such technology has urgent requirements for storage reduction and efficient use of the transmission channel bandwidth. Such requirements can be achieved efficiently via developing accurate medical image compression techniques. For accurate diagnosis, lossless compression methods are recommended. However, the tradeoff between the compression ratio (CR) and the preservation of the image quality is still challenging. On the other hand, the advantages of the convolutional neural networks inspired this work to design a novel proposed system for dermoscopic image compression based on the integration of the compression direction-ConvNet (CD-ConvNet) and decompression direction-ConvNet (DD-ConvNet) with discrete cosine transform (DCT) and Huffman coding, called DermCompressNet. To reconstruct a high-quality image at the receiver, the inverse processes using the DD-ConvNet network were followed. The proposed system was evaluated by measuring several image quality metrics, namely the mean square error (MSE), peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM), along with compression quality metrics, namely CR, and computational time (CT). The experimental results achieved 34.6 dB, 2.5, 0.85, and 56% of PSNR, MSE, SSIM, and CR, respectively. A comparison studies with the JPEG and state-of-the-art methods proved the superiority of the proposed system, showing23%, 16%, 4.3%, and 1.8% improvements in the PSNR, MSE, SSIM, and CR, respectively, compared to the JPEG.

LeCo: Lightweight Compression via Learning Serial Correlations

Lightweight data compression is a key technique that allows column stores to exhibit superior performance for analytical queries. Despite a comprehensive study on dictionary-based encodings to approach Shannon's entropy, few prior works have systematically exploited the serial correlation in a column for compression. In this paper, we propose LeCo (i.e., Learned Compression), a framework that uses machine learning to remove the serial redundancy in a value sequence automatically to achieve an outstanding compression ratio and decompression performance. LeCo presents a general approach to this end, making existing algorithms such as Frame-of-Reference (FOR), Delta Encoding, and Run-Length Encoding (RLE) special cases under our framework. Our microbenchmark with three synthetic and eight real-world data sets shows that a prototype of LeCo achieves a Pareto improvement on both compression ratio and random access speed over the existing solutions. When integrating LeCo into widely-used applications, we observe up to 5.2× speed up in a data analytical query in the Arrow columnar execution engine, and a 16% increase in RocksDB's throughput.

Information about the book by Joseph Dubrovkin “Data Compression IN Spectroscopy”, Cambridge Scholars Publishing. 2022, 355 PP. ISBN (10): 1-5275-8620-0 ISBN (13)

Widespread use of instrumental methods of analysis, in particular analytical spectroscopy, in recent years has led to increasing volume of information about the objects being studied. Experimental results are stored on a computer in the form of multidimensional digital data arrays that need to be processed in a special way, for example, to remove noise, visualize, analyze, and compress for efficient storage. For these purposes, there are theoretical methods and algorithms of data compression. At the same time compression (decompression) algorithms must ensure minimum distortion of the original signals for processing hyperspectral data. Large volume of data implies a comprehensive analysis of information. When working with multidimensional arrays, it is important to use special methods and technologies for processing, analyzing and interpreting hyperspectral data. This article provides information about the monograph “Data Compression in Spectroscopy” by Joseph Dubrovkin. The book was published by Cambridge Scholars Publishing. The book was written by an expert with extensive experience in the field of multivariate data analysis and chemometrics. The book consists of a preface, information about the structure of the book, a list of abbreviations and symbols, an introduction for each of 4 chapters, 8 appendices, a list of references and a subject index. A large number of examples and exercises are illustrated with MATLAB programs, and bibliographic tables clearly demonstrate the use of compression methods in industrial and research laboratories. The material in the book is discussed chapter by chapter. This modern monograph on data compression in spectroscopy will be useful as a teaching aid for students and teaching staff, as well as for specialists in analytical laboratories.

Abstract 189: How Quickly Does Arterial Blood Pressure Recover After Pauses in Mechanical Chest Compressions in Humans With Out-of-Hospital Cardiac Arrest?

Introduction: Conventional wisdom says that arterial blood pressure (ABP) recovers slowly after interrupting CPR chest compressions (CCs). An oft-cited animal study with manual compressions and 15:2 compression:ventilation ratio showed that all 15 compressions were needed to restore pre-pause invasive pressure, implying that even short pauses substantially reduce CPR blood flow. Whether this holds true for humans receiving mechanical chest compressions is unknown. We measured recovery of ABP after pauses in mechanical compressions by LUCAS in patients treated for out-of-hospital cardiac arrest (OHCA). Methods: We analyzed data from a subset of patients enrolled in both groups of a prospective, randomized LUCAS 2 Active Decompression trial performed by an anesthesiologist-staffed rapid response car program in Oslo, Norway from 2016 to 2018. Only patients with an ABP signal and at least one CC pause longer than 2 sec were included. All CCs were mechanical. Arterial cannulation was verified, and CC pauses were identified by physician review of the patient care record and physiological signals. Pauses were excluded if ROSC or a LUCAS “soft start” occurred or epinephrine had been applied within 5 minutes of the pause. Peak compression (cAP), mean (mAP), and mean decompression arterial blood pressures (dAP) for 10 CCs before and after each pause were measured with custom MATLAB code. Results: We included 56 patients with a total of 148 pauses (mean pause duration: 10.63 ± 3.05 sec). Pre-pause mean cAP, mAP, and dAP were 95.6 ± 0.3, 49.9 ± 0.1 and 30.5 ± 0.1 mmHg. Post-pause dAP for each compression was not significantly different than pre-pause dAP (Fig. 1A,B). cAP and mAP recovered to >90% of pre-pause pressure within 2 CCs. Pause duration was not predictive of post-pause pressures (Fig. 1C). Conclusions: ABP generated by mechanical CPR recovered quickly after pauses. Consequently, short pauses (such as pauses for 30:2 CPR) may pose less hemodynamic concern than previously thought.

Abstract 385: Comparison of Rapid Implementation of Automated Head and Thorax Elevation With Active Compression Decompression Cardiopulmonary Resuscitation and an Impedance Threshold Device versus National Outcome Data: A Prospective Observational Study

Background: Pre-clinical work has shown automated head and thorax elevation with active compression decompression CPR and an impedance threshold device (AHUP-CPR) improved 24-hour neurological survival. Propensity matched studies of out-of-hospital-cardiac arrest (OHCA) patients have shown a time dependent association between survival to hospital discharge (SHD) and application of AHUP CPR.Hypothesis: Rapid application of AHUP CPR for OHCA will have a higher incidence of SHD compared to national data. Methods: This is an ongoing prospective observational study of patients from an AHUP-CPR registry of non-traumatic adult OHCA receiving AHUP CPR as part of first responder care. OHCA patients from pre-hospital sites that consistently implemented rapid AHUP CPR were included. Patient demographics, arrest characteristics, ROSC, SHD, and SHD with favorable neurological function, defined as a Cerebral Performance Category score of 1 or 2, were recorded. Data are reported as means ± SD. Combined 2021 and 2022 data from the Cardiac Arrest Registry to Enhance Survival (CARES) were used for comparison of the primary study endpoint, SHD, and secondary endpoints. Chi-square testing was used to compare outcomes. Results: Of 11 registry sites, 7 (US States AL, FL, IL, MN, OK, TN, VA) met inclusion criteria and 1457 patients from January 2020 to June 2023 were enrolled. Average age was 65 ± 16.4 years, 61.8% (901) were male, and 71% (1035) received AHUP CPR. Average time (mm:ss) from 9-1-1 call to start of CPR was 9:52 (6:47) and time to start of AHUP CPR was 12:59 (5:51). Presenting rhythms were: asystole 47.8% (696), PEA 25.8% (376), and VF 20.5% (298). For all AHUP patients, SHD was 13.6% (198) versus 9.2% (26316/286525) for CARES (p < 0.0001). SHD with favorable neurological function was higher for AHUP (10.3% (150) versus 7.3% for CARES (20889), p < 0.0001) and ROSC (29.4% (428) vs 27.0% (77301), p = 0.039). Conclusions: In an unadjusted analysis, the incidence of SHD, ROSC, and neurologically favorable SHD with rapid implementation of AHUP-CPR was higher than nationally reported data. These results support prior findings of a time sensitive association between survival and time to AHUP-CPR. Additional analyses and study are needed for confirmation.

Refactoring BZIP2 on the new‐generation sunway supercomputer

AbstractHigh‐performance computing is progressively assuming a fundamental role in advancing scientific research and engineering domains. However, the ever‐expanding scales of scientific simulations pose challenges for efficient data I/O and storage. The data compression technology has garnered significant attention as a solution to reduce data transmission and storage costs while enhancing performance. In particular, the BZIP2 lossless compression algorithm has been widely used due to its exceptional compression ratio, moderate compression speed, high reliability, and open‐source nature. This paper focuses on the design and realization of a parallelized BZIP2 algorithm tailored for deployment on the New‐Generation Sunway supercomputing platform. By leveraging the unique cache patterns of the New‐Generation Sunway processor, we propose the highly tuned multi‐threading and multi‐node implementations of the BZIP2 applications for different scenarios. Moreover, we also propose the efficient BZIP2 libraries based on the management processing element and computing processing element which support the commonly used high‐level (de)compression interfaces. The test results indicate that the our multi‐threading implementation achieves maximum speedup of 23.09 (8.57) in decompression(compression) compared to the sequential implementation. Furthermore, the multi‐node implementation achieves 50.81% (26.35%) parallel efficiency and peak performance of 16.6 GB/s (52.8 GB/s) for compression(decompression) when scaling up to 2048 processes.

Elastic–plastic intermittent re-arrangements of frictionless, soft granular matter under very slow isotropic deformations

How do soft granular materials (or dense amorphous systems) respond to externally applied deformations at different rates and for different system sizes? This long-standing question was intensively studied for shear deformations but only more recently for isotropic deformations, like compression–decompression cycles. For moderate strain rates, in the solid-like state, above jamming, the system appears to evolve more or less smoothly in time/strain, whereas for slow enough deformations, the material flips intermittently between the elastic, reversible base state and plastic, dynamic “events.” Only during the latter events, the microstructure re-arranges irreversibly. The reversible base state involves both affine and non-affine deformations, while the events are purely non-affine. The system size and rate dependence of the events are studied, providing reference data for comparison in future studies evaluating materials like hydrogel particles.

Retinal Biometric for Securing JPEG-Codec Hardware IP Core for CE Systems

Fake/pirated hardware intellectual property (IP) cores integrated in consumer electronics systems can cause reliability hazards and jeopardize the safety of end consumer. Robust detection and isolation of such pirated IP cores before integration is therefore crucial. Existing approaches do not provide robust security against replication or evasion of IP detection resulting into higher probability of coincidence (Pc) and lesser tamper tolerance (TT) compared to the proposed approach. Further, compared to existing biometric based hardware security methodologies, the proposed approach provides more distinctive features and does not require image enhancement; while compared to non-biometric based hardware security methodologies, the proposed approach provides natural uniqueness and non-replicability of features through retinal image. This paper presents a novel retina biometric-based hardware security methodology for securing JPEG compression-decompression (CODEC) hardware IP core against IP piracy. The proposed result reports the following: (a) variation in Pc metric for different sizes of embedded retinal signature and different retinal images (b) variation of TT for proposed approach for different retinal images (c) security comparison of proposed approach with facial, fingerprint biometric and encrypted digital signature-based hardware security approaches (d) Pc-design cost tradeoff for proposed retinal biometric approach. Results indicate enhancement in security at zero design overhead.

Comparison of percutaneous balloon compression and microvascular decompression in the treatment of trigeminal neuralgia.

To compare the effect of percutaneous balloon compression (PBC) and microvascular decompression (MVD) in the treatment of trigeminal neuralgia (TN). Data of 98 patients with TN, admitted to Chenzhou First People's Hospital from May 2020 to May 2022, were retrospectively collected. Patients were divided into two groups based on the surgical method. A total of 53 patients treated with PBC comprised the PBC-group and 45 patients treated with MVD comprised the MVD-group. The immediate pain relief, long-term efficacy, surgical complications, and masticatory muscle strength of the two groups were compared and analyzed. There was no significant difference in the immediate pain relief and long-term e efficacy, between the two groups (P>0.05). Complication rate in the PBC-group was significantly lower than that in the MVD-group (3.77% vs 17.78%, P<0.05). Medical records within 14 days after the operation showed that the incidence of facial numbness and masticatory muscle weakness in the PBC-group were 37.74% and 28.30% respectively, significantly higher than those in MVD-group (4.44% and 2.22%) (P<0.05). These symptoms gradually improved three months after the surgery, and were almost completely resolved after six months. Compared with MVD, PBC has the same effect in the treatment of TN. PBC is a minimally invasive, safe, and effective method with a low complication rate. Although masticatory muscle strength is slightly impacted by PBC, it gradually recovers within six months after the operation.

A Comparative Review of the Outcome Following MVD and PBC in Patients with Trigeminal Neuralgia.

This study aims to systematically review the treatment outcomes of percutaneous balloon compression (PBC) and microvascular decompression (MVD) in patients with trigeminal neuralgia. A systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guideline was performed using PubMed, Embase, and Cochrane Central Registry of Controlled Trials databases. Only those articles with more than 5 years' follow-up length were included in this investigation. To uniformly assess the postoperative outcome, we defined pain relief as totally pain free, while the postoperative hospitalization and last follow-up period were defined as early and long term, respectively. The facial numbness was quantified with Barrow Neurological Institute Pain Intensity Score (BNI). After database searching and screening, 7,797 cases were finally included according to the criteria. The early pain relief rates were 94.1% (1,551/1,649) and 89.9% (4,962/5,482) following PBC and MVD (odds ratio [OR] = 0.603; p < 0.05), while the long-term rates were 58.1% (921/1,566) and 74.9% (4,549/6,074; OR = 2.089; p < 0.05), respectively. Although a significant higher facial numbness occurred in the PBC group in the early stage, it was mostly diminished 5 years later compared with the MVD group. At long-term follow-up, hypoacusis and facial palsy occurred more often in the MVD group (p < 0.05). Both MVD and PBC provide a satisfactory outcome for the patients in the long term. As a simple, safe, and reliable technique, PBC should be considered as a viable alternative.

A Novel Low-Power Compression Scheme for Systolic Array-Based Deep Learning Accelerators

The proliferation of deep learning algorithms has catalyzed their utilization to solve a multitude of real-world problems. Algorithms such as deep neural networks (DNNs) are compute- and power-intensive, thereby accentuating the development of hardware platforms like DNN inference accelerators. However, inference execution of large DNNs in resource-constrained environments induces energy bottlenecks in these accelerators. Since large DNNs consist of hundreds of millions of trained parameters, accessing them from the accelerator memory incurs substantial energy. To address this challenge, we propose HardCompress, which, to the best of our knowledge, is the first low-power solution that uses traditional compression strategies pertaining to commercial DNN accelerators in resource-constrained IoT edge devices. The three-step approach involves hardware-based post-quantization trimming of weights, followed by their dictionary-based compression and subsequent decompression by a low-power hardware engine during inference in the accelerator. We evaluate the proposed solution on lightweight networks trained on the MNIST dataset, the compact model trained on the CIFAR-10 dataset, and large DNNs trained on the ImageNet dataset. Performance of HardCompress at different quantization levels has been analyzed. Furthermore, to quantify the effectiveness of the proposed solution, an energy framework that contrasts the DRAM energies of the original and HardCompressed models has been developed. Finally, a fault injection framework which compares the fault resilience of the original model with its HardCompressed counterpart is also proposed. Our results exhibit that HardCompress, without any performance degradation in large DNNs, furnishes a maximum compression of 99.27%, equivalent to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$137\times $ </tex-math></inline-formula> reduction in memory footprint and 0.07 J for 8-bit quantization in the systolic array-based DNN accelerator. Furthermore, our proposed low-power decompression engine incurs an area overhead of only 0.02%; thus, enabling HardCompress’ utilization in resource-constrained environments.

Carboxymethyl Cellulose-Based Composite Polymer Hydrogels Cross-Linked with Epichlorohydrin and Application for Cu(II) Removal

In this work, CMC-ECH/PAM hydrogels with perfect compressive strength were prepared by epichlorohydrin (ECH) cross-linking sodium carboxymethyl cellulose (CMC) grafted with polyacrylamide (PAM) to remove the copper ion (Cu2+) from water. The chemical and structural morphology of the product was confirmed by Fourier transform infrared spectroscopy and scanning electron microscopy, respectively. The compressive strength was measured by successive compression–decompression cycles and up to 70 kPa. Adsorption was studied at different pH values 1–5, adsorbent doses (0.2–8) g/L, and initial Cu2+ concentrations from (50 to 400) mg/L to obtain the optimum conditions for maximum removal of Cu2+. The experimental adsorption data were well fitted by the pseudo-second-order kinetics model, Langmuir isotherm model, and Freundlich isotherm model. The maximum adsorption capacity calculated by the Langmuir model was 75.930 mg/g at 338.15 K. In addition, the negative ΔG° and positive ΔH° also indicated that the adsorption of Cu2+ is a spontaneous, endothermic, and chemical process. The adsorption mechanism was further verified by X-ray photoelectron spectroscopy analysis. CMC-ECH/PAM hydrogels also exhibited excellent reusability after five adsorption–desorption cycles, maintaining about 80% Cu2+ removal efficiency.

Lightweight Video Super-Resolution for Compressed Video

Video compression technology for Ultra-High Definition (UHD) and 8K UHD video has been established and is being widely adopted by major broadcasting companies and video content providers, allowing them to produce high-quality videos that meet the demands of today’s consumers. However, high-resolution video content broadcasting is not an easy problem to be resolved in the near future due to limited resources in network bandwidth and data storage. An alternative solution to overcome the challenges of broadcasting high-resolution video content is to downsample UHD or 8K video at the transmission side using existing infrastructure, and then utilizing Video Super-Resolution (VSR) technology at the receiving end to recover the original quality of the video content. Current deep learning-based methods for Video Super-Resolution (VSR) fail to consider the fact that the delivered video to viewers goes through a compression and decompression process, which can introduce additional distortion and loss of information. Therefore, it is crucial to develop VSR methods that are specifically designed to work with the compression–decompression pipeline. In general, various information in the compressed video is not utilized enough to realize the VSR model. This research proposes a highly efficient VSR network making use of data from decompressed video such as frame type, Group of Pictures (GOP), macroblock type and motion vector. The proposed Convolutional Neural Network (CNN)-based lightweight VSR model is suitable for real-time video services. The performance of the model is extensively evaluated through a series of experiments, demonstrating its effectiveness and applicability in practical scenarios.

Photoplethysmography Data Reduction Using Truncated Singular Value Decomposition and Internet of Things Computing

Biometric-based identity authentication is integral to modern-day technologies. From smart phones, personal computers, and tablets to security checkpoints, they all utilize a form of identity check based on methods such as face recognition and fingerprint-verification. Photoplethysmography (PPG) is another form of biometric-based authentication that has recently been gaining momentum, because it is effective and easy to implement. This paper considers a cloud-based system model for PPG-authentication, where the PPG signals of various individuals are collected with distributed sensors and communicated to the cloud for authentication. Such a model incursarge signal traffic, especially in crowded places such as airport security checkpoints. This motivates the need for a compression–decompression scheme (or a Codec for short). The Codec is required to reduce the data traffic by compressing each PPG signal before it is communicated, i.e., encoding the signal right after it comes off the sensor and before it is sent to the cloud to be reconstructed (i.e., decoded). Therefore, the Codec has two system requirements to meet: (i) produce high-fidelity signal reconstruction; and (ii) have a computationallyightweight encoder. Both requirements are met by the Codec proposed in this paper, which is designed using truncated singular value decomposition (T-SVD). The proposed Codec is developed and tested using a publicly available dataset of PPG signals collected from multiple individuals, namely the CapnoBase dataset. It is shown to achieve a 95% compression ratio and a 99% coefficient of determination. This means that the Codec is capable of delivering on the first requirement, high-fidelity reconstruction, while producing highly compressed signals. Those compressed signals do not require heavy computations to be produced as well. An implementation on a single-board computer is attempted for the encoder, showing that the encoder can average 300 milliseconds per signal on a Raspberry Pi 3. This is enough time to encode a PPG signal prior to transmission to the cloud.

Linearization Weight Compression and In-Situ Hardware-Based Decompression for Attention-Based Neural Machine Translation

As recent machine translation models are mostly based on the attention-based neural machine translation (NMT), many well-known models such as Transformer or bidirectional encoder representations from Transformers (BERT) have been proposed. Along with algorithmic advancements, hardware acceleration methods for those attention-based neural machine translation models have also been introduced. However, the size of the parameters for attention-based NMT is also becoming larger to guarantee the satisfactory machine translation quality. Among various weights, linearization weights ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$W^{Q}$ </tex-math></inline-formula> , <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$W^{K}$ </tex-math></inline-formula> , <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$W^{V}$ </tex-math></inline-formula> , and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$W^{O}$ </tex-math></inline-formula> ) account for a non-negligible portion (by up to 30%) among the entire parameters in the modern NMT models. In this paper, we propose a method for linearization weight compression and near-memory hardware decoder for fast and in-situ weight decompression. Our weight compression method exploits the fixed-point quantization along with Huffman coding which is selectively applied depending on the weight value distribution. Our hardware decoder decompresses the Huffman-coded weights near-memory to minimize the weight decoding latency. Our compression method shows 4.9–10.0 compression ratio with small NMT score drops across the five widely used attention-based NMT models (Transformer, Transformer-XL-base, Transformer-XL-large, BERT-base, and BERT-large). In addition, due to the reduced linearization weight size, our proposed method with near-memory decoding enables multi-head attention (MHA) execution latency reduction by 11.8%, on average, as compared to the baseline when considering the weight loading and initialization. In terms of the memory data transfer energy consumption, our proposed method leads to a memory energy saving of 16.1%, on average, as compared to the baseline.

High-Performance Hydrogels via Alternate Compression–Decompression

Novel bioapplications of hydrogels draw huge attention to the development of strong and tough hydrogels that are easily obtainable with less chemical additions. Here, we demonstrate that pressure, a basic thermodynamic quantity, could enable poly(vinyl alcohol) (PVA) to effectively form a gel. The resulting hydrogels by alternate compression–decompression (ACD) have tunable superior mechanical properties compared with conventional freeze-thawed (FT) PVA hydrogels. The microstructures of the ACD hydrogels under varying parameters, including compression rate, cycles, and holding time, reveal that either a slow compression rate or cyclic compressions favor the physical cross-linking of such single polymer networks. The mechanical results display a wide range of ultimate tensile strength, tensile strain, and maximum compressive strength of ∼0.3 to 2.5 MPa, ∼200 to 550%, and ∼3 to 7 MPa, respectively. Moreover, the load resistance of such hydrogels can be further trained by low-cyclic strain hardening to a maximum compressive strength of ∼50 MPa, followed by a desirable recovery. Upon cyclic compression, the ACD hydrogels exhibit consistent energy dissipation behaviors. More importantly, the effect of modulation of pressure on the hydrogel’s mechanical properties is very likely universal for other hydrogel systems due to the basic mechanism of pressure-induced gelation for the polymers discussed.

Abstract 11244: Active Decompression During Device Assisted Elevation of the Head and Thorax, an Impedance Threshold Device, and Active Compression Decompression Cardiopulmonary Resuscitation is Needed for Sustained Hemodynamic Benefit in a Porcine Model of Prolonged Cardiac Arrest

Introduction: Active compression decompression cardiopulmonary resuscitation (CPR), an impedance threshold device (ITD), and automated device-assisted controlled head up positioning, collectively termed AHUP CPR, work synergistically to lower intracranial pressure, augment venous return to the heart and significantly improve neurological survival versus conventional (C)-CPR in animals and patients in cardiac arrest. The current study focused on the role of active decompression (AD) during prolonged AHUP CPR, since AD can be difficult to perform manually for an extended period. Hypothesis: Full AD is needed to optimize hemodynamics during prolonged AHUP CPR Methods: Six female and male farm pigs (~40 kg) were sedated, intubated, anesthetized, and monitored. After 8 min of untreated ventricular fibrillation followed by ≥ 30 min of AHUP CPR with an automated device, AHUP CPR was continued except that AD was intentionally reduced from an average of 3.33 ± 0.37 cm of lift to 0 cm lift above the resting neutral position as this is the amount of decompression that occurs with C-CPR and passive chest wall recoil. After 1 min of AHUP CPR with 0 cm lift, AD was progressively increased back to 3 cm of lift over 3 min. Hemodynamics (mean ± SD) were compared using Student’s t test, where p&lt;0.05 was considered statistically significant. Results Coronary perfusion pressure (CorPP) (p&lt;0.005), cerebral perfusion pressure (CerPP) (p&lt;0.001), and ETCO2 (p&lt;0.0001) significantly decreased after AD was discontinued (Fig.1). With restoration of AD to 3 cm, CorPP (p&lt;0.009) and ETCO2 (p&lt;0.001) significantly improved and CerPP trended higher (p=0.084). Conclusion During prolonged (≥ 30 min) AHUP CPR in pigs, active lift of the chest wall is essential to maintain stable hemodynamics. Hemodynamics deteriorated rapidly with intentional loss of AD, but could be partially restored with resumption of AD.

Pressure Engineering Promising Transparent Oxides with Large Conductivity Enhancement and Strong Thermal Stability.

Transparent conducting oxides (TCO) with high electrical conductivity and high visible light transparency are desired for a wide range of high-impact engineering. Yet, usually, a compromise must be made between conductivity and transparency, limiting the practical application of a TCO to the next level. Furthermore, TCO performance is highly sensitive to composition, so conventional synthesis methods, such as chemical doping, cannot unravel the mysteries of the quantitative structure-performance relationship. Thus, improving the fundamental understanding or creating materials-by-design has limited success. Here, a strategy is proposed to modulate the lattice and electronic and optical properties precisely by applying pressure on a TCO. Strikingly, after compression-decompression treatment on the indium titanium oxides (ITiO), a highly transparent and metastable phase with two orders of magnitude enhancement in conductivity is synthesized from an irreversible phase transition. Moreover, this phase possesses previously unattainable filter efficiency on hazardous blue light up to 600°C, providing potential for healthcare-related applications with strong thermal stability up to 200°C. These results demonstrate that pressure engineering is a clean and effective tool for tailoring functional materials that are not achievable by other means, providing an exciting alternative property-tuning dimension in materials science.