搜索结果：code

A comprehensive transthoracic echocardiogram involves the assessment of over 70 parameters, placing a substantial burden on sonographers and physicians for manual annotation with considerable inter-observer variability. Prior open-source segmentation models have largely addressed 2D B-mode ventricular function, leaving a gap in the spectral Doppler and atrial measurements required for valvular and diastolic assessment such as velocity-time integral (VTI) and atrial chamber size. In this retrospective multi-cohort study, we developed EchoNet-Segmentation, comprehensive task-specific deep learning segmentation models for left and right atrial area and VTI Doppler measurements. Training used 186,712 sonographer-annotated images from 93,978 studies (56,855 patients) at Cedars-Sinai Medical Center (CSMC). Performance was evaluated on a held-out CSMC test set, a CSMC temporal split, an external Kaiser Permanente Northern California cohort, and the public MIMIC-Echo dataset. On the CSMC held-out test set, our AI models showed strong agreement with sonographer measurements, with R² of 0.817-0.882 and mean absolute error (MAE) of 1.13-3.80 cm for automated VTI measurements, and R² of 0.675-0.747 and MAE of 2.48-2.52 cm² for left and right atrial area segmentation. Performance was consistently confirmed on the CSMC temporal split (VTI: R² 0.606-0.866, atrial area: R² 0.694-0.705) and on the KPNC external cohort (VTI: R² 0.575-0.859, atrial area: R² 0.803-0.876), on the MIMIC-Echo dataset. Robustness was demonstrated on a different vendor's machines and across subgroups. EchoNet-Segmentation outperformed an open-source medical image foundation model with bounding-box, point prompt configurations on R², MAE, and Dice score on both held-out test dataset and MIMIC apical four-chamber data. EchoNet-Segmentation is the first open-source framework that delivers accurate, generalizable automated measurement across several key routine echocardiographic parameters, supporting end-to-end automation of clinically important echocardiographic assessments. Public release of model weights, code, and demonstration tools can facilitate reproducibility, research use and clinical deployment. Funding Statement: This work was supported by NIH NHLBI grants R00HL157421, R01HL173526, and R01HL173487 to D.O. Evidence before this study: We searched PubMed and arXiv from database on April 1, 2026, for studies of deep learning-based segmentation of echocardiographic images, using the terms ("echocardiography" OR "echocardiogram") AND ("deep learning" OR "artificial intelligence") AND ("segmentation" OR "measurement"). Prior work has demonstrated automated segmentation of cardiac chambers and left ventricular ejection fraction estimation, and a small number of studies have reported deep learning models for velocity-time integral (VTI) or atrial size measurement. However, openly available models and code remain largely restricted to left ventricular structures, ejection fraction, and wall thickness, and commercial tools remain proprietary. To our knowledge, no open-source framework has comprehensively addressed VTI measurements across multiple Doppler views together with atrial chamber size in a single, reproducible toolkit, and existing models have not been systematically benchmarked against general-purpose medical-image foundation models on echocardiographic tasks.Added value of this study: We developed and validated EchoNet-Segmentation, a suite of task-specific deep learning models for several clinically important echocardiographic parameters: left and right atrial area and five VTI measurements (aortic valve, mitral valve, left ventricular outflow tract, right ventricular outflow tract, and pulmonary valve). The models were trained on the largest real-world collection of sonographer-annotated echocardiograms reported to date (186,712 images from 56,855 patients) in an academic center in the United States and showed strong agreement with sonographer measurements on a held-out internal test set, a temporal split cohort, an external cohort from a different health system, and a publicly available cohort recorded on a different vendor's ultrasound machines. EchoNet-Segmentation outperformed the publicly released medical-image foundation model (MedSAM2) on cardiac chamber segmentation across both internal and public dataset benchmarks. All model weights, training and inference code, demonstration tools, and the manual segmentation masks used for the public benchmark are openly released.Implications of all the available evidence: EchoNet-Segmentation enables end-to-end automation of routine transthoracic echocardiographic measurements with previously released open-source models. By openly releasing model weights, training code, and benchmark data, this work provides a reproducible foundation that the broader research and clinical community can build on, fine-tune for specific populations or imaging protocols, and integrate into clinical workflows. Prospective validation and randomized studies will be needed to define the impact of automated measurement on diagnostic accuracy, workflow efficiency, and clinical outcomes.

Throughout the mammalian cortex, populations of neurons must work together to enact behaviors. While population recordings in motor cortex have revealed many aspects of when neurons fire during behaviors, limitations in causal experiments have made it difficult to identify which features of neural activity directly drive movements and which do not. Here, we explicitly test the principles of neural coding using high temporal precision multiphoton holographic optogenetics in the motor cortex. We show that activation of a small number (50-75) of Layer 2/3 excitatory neurons in the motor cortex is sufficient to drive movements. The efficacy of stimulation-driven movement depends on the state of the local circuit, and, to a lesser extent, the identity of which neurons are stimulated. We test whether evoked activity is acting through a 'rate code' or a 'timing code' by holding the firing rate and cell identities constant while varying the millisecond precise timing of activation. We find an unexpected and strong dependence on inter-cell synchrony when evoking movements. This neural synchrony recruits distinct patterns of recurrent excitation and inhibition. These findings provide evidence that the timing code, more so than the rate code, drives motor output.

Manual ICD-10 coding of drug contraindications is resource-intensive and difficult to scale. Large language models (LLMs) have shown early promise for clinical information extraction, but their reliability for structured pharmacological coding relative to validated expert databases remains unknown. To assess the concordance between LLM-based automated ICD-10 coding and Thésorimed drug database1 for antihypertensive drug contraindications. Contraindications were extracted from Summary of Product Characteristics free-text documents for 301 antihypertensive drugs using a retrieval-augmented generation pipeline and aligned against Thésorimed ICD-10 codes. Agreement was assessed using a binary kappa (κ1, entity presence) and a hierarchical weighted kappa (κ2, ICD-10 code concordance) on 5,074 entities. Entity-level agreement was slight (κ1 = -0.30), reflecting structural differences in source coverage. Code-level agreement among matched pairs was substantial (κ2 = 0.70, Po = 0.871), with 76.9% identical codes. When both sources identify the same contraindication, LLM-based coding shows substantial concordance with expert-curated ICD-10 assignments. However, low entity-level agreement highlights important differences in coverage and representation between sources.

Socioeconomic status and sex each independently influence stroke outcomes. However, their intersectional effects remain unclear. Accordingly, we sought to assess whether the association between socioeconomic status and outcomes in stroke varied by sex. We conducted a retrospective cohort study using the National Inpatient Sample (2016-2021). Patients with a primary diagnosis of acute ischemic stroke, intracerebral hemorrhage, or subarachnoid hemorrhage were analyzed in separate cohorts. The primary exposure was socioeconomic status (ZIP code income quartile). Multivariable logistic regression was used to assess the effect of socioeconomic status on stroke outcomes. Significant first-order sex-income interactions were observed across all stroke subtypes; accordingly, we built sex-stratified models. A total of 2&#x2009;836&#x2009;254 acute ischemic stroke, 406 550 intracerebral hemorrhage, and 134&#x2009;510 subarachnoid hemorrhage admissions were included. In male patients, patients from the lowest quartile of ZIP code income quartile had significantly increased odds of inpatient mortality across acute ischemic stroke (odds ratio [OR], 1.11 [95% CI, 1.03-1.21], P=0.009), intracerebral hemorrhage (OR, 1.15 [95% CI, 1.05-1.25], P=0.002), and subarachnoid hemorrhage (OR, 1.46 [95% CI, 1.22-1.74], P<0.001). In female patients, ZIP code income quartile was not associated with inpatient mortality for any stroke subtype: acute ischemic stroke (OR, 1.04 [95% CI, 0.96-1.12], P=0.361), intracerebral hemorrhage (OR, 1.09 [95% CI, 1.00-1.20], P=0.052), and subarachnoid hemorrhage (OR, 1.04 [95% CI, 0.90-1.20], P=0.601). Socioeconomic disadvantage was associated with worse outcomes in stroke among male but not female patients. These findings highlight a sex-specific disparity in stroke outcomes and underscore the need for public health initiatives targeting male patients experiencing socioeconomic disadvantage.

Large language models (LLMs) are increasingly explored for qualitative analysis, but the effect of workflow design on thematic fidelity remains unclear. This study evaluated a structured human-AI collaboration framework using Claude Opus 4.6 to analyze 16 interview transcripts from patients with chronic obstructive pulmonary disease participating in a pulmonary telerehabilitation program. The workflow included code extraction, code combination, and theme generation, and was tested using hierarchical and direct strategies. AI-generated themes were compared with human-derived themes using sentence-t5-xxl embeddings and cosine similarity, with theme alignment performed using Hungarian and greedy matching. Output volume varied substantially across strategies, ranging from 53 to 357 codes and 11 to 17 themes. Direct grouping (average cosine similarity 0.891) and L3 grouping (0.890) achieved the highest similarity to human-generated themes. These findings suggest that grouping-based workflows can preserve key information, reduce redundancy, and improve thematic generation in LLM-assisted qualitative analysis.

Since the outbreak of the COVID-19 pandemic in 2020, people in Slovenia have faced a series of overlapping crises at both global and national levels, including the Russo-Ukrainian conflict, rising economic costs, and major natural disasters. Although extensive quantitative research has examined pandemic-related mental health outcomes, comparatively little is known about how individuals subjectively perceive multiple overlapping crises, given the cumulative and interactive nature of contemporary crises - situations in which crises across multiple systems become causally entangled and mutually amplify their effects. This study aimed to explore the perceived psychological consequences of multiple overlapping crises among adults in Slovenia, guided by the polycrisis framework. Using a qualitative research design based on Constructivist Grounded Theory, we conducted semi-structured, in-depth interviews between June and November 2023 with 24 participants (16 women, 8 men; mean age 49 years, range 21-76). Interviews were audio-recorded, manually transcribed, and analyzed using ATLAS.ti 24 software. Analysis involved initial and focused coding, constant comparison, and analytic memo-writing, continuing until theoretical saturation was reached. A total of 274 quotations were coded into 16 final codes. Three analytical categories of perceived psychological consequences emerged: duration (short-term vs. long-term), valence (positive vs. negative), and level of impact (individual vs. societal). Short-term consequences were predominantly individual-level experiences tied to the acute pandemic period, including negative outcomes (e.g., stress, mental health decline, social polarization) and positive outcomes (e.g., increased time with close ones, time in nature, reduced pressure). Positive long-term consequences included closer relationships and increased gratitude; negative ones included a reduced sense of safety, distrust in authorities, weaker social ties, and a perceived decline in social cohesion. Based on the analytical findings, we developed the Fortress Wall Model of Psychological Adaptation to Crisis, which conceptualizes the experience of compounding crises as a dynamic, phased psychological process. The findings underscore the need to address people's emotional experiences during acute crisis phases and to maintain monitoring and support of psychological well-being beyond the acute period, as negative consequences - particularly heightened uncertainty, distrust, and a diminished sense of safety - persist long after crises subside and are intensified by subsequent stressors.

Research reproducibility - the ability of others to independently verify scientific findings by following the same methods and data - is a fundamental aspect of research integrity. When this fails, trust in evidence is undermined and research resources are wasted. Reproducibility measures describe the standards that define reproducible research, and reproducibility interventions are the actions used to improve those standards. There is little agreement on which practices and initiatives should be prioritised for implementation in practice. This study aimed to establish expert consensus on the key measures and interventions that should be prioritised to strengthen research reproducibility. We conducted a Delphi consensus study as part of the EU Horizon Europe iRISE project. Experts from five stakeholder groups (researchers, editors, publishers, funders, and policymakers) evaluated reproducibility measures and interventions identified through prior literature mapping across two online survey rounds, followed by a final virtual consensus panel. Items were rated on a 10-point Likert scale, with consensus defined a priori as agreement by at least 70% of panellists assigning high-priority scores, corresponding to scores of eight to ten on the Likert scale. Seventy-three panellists from 34 countries participated in the first round, with high retention rates in the subsequent round. Consensus was achieved on eight reproducibility measures and six reproducibility interventions in the first round. Prioritised measures included methodological quality, reporting quality, code and data availability and reuse, computational reproducibility, transparency of research plan, reproducible workflow practices, trial registration, and materials availability and reuse. Prioritised interventions included data management training, data quality checks/feedback, statistical training, data sharing policy/guidelines, protocol/trial registration, and reproducible code/analysis training. The prioritised measures and interventions provide a structured foundation for improving research reproducibility across disciplines. The findings can inform the development of institutional training curricula in data management and statistical methods, policies supporting data sharing and trial registration, and future empirical evaluation of these practices across research contexts.

Hospitals often struggle to reuse clinically relevant data because specialty information remains tied to application-specific structures. This study examines how an existing ambulatory blood pressure monitoring (ABPM) module can be translated into an openEHR-based template and prototype in a university hospital setting. An exploratory implementation case study combined workflow analysis, archetype review, gap analysis, template design, and low-code prototyping. The resulting template represented repeated blood pressure observations, interval-based summaries, encounter context, pulse-related parameters, and interpretation within one coherent structure. The case showed that adequate ABPM representation required a constrained combination of several archetypes rather than reuse of the blood pressure archetype alone. Prototyping further demonstrated that low-code implementation can serve as an early validation step for template quality. Beyond the ABPM use case, the study contributes a practical modeling pattern for translating specialty modules into constrained openEHR templates.

BackgroundPatient-led melanoma surveillance using mobile dermoscopy and teledermatology may support earlier detection and improve access to dermatology care for underserved populations.MethodsWe conducted semi-structured interviews with Melanoma Self Surveillance (MEL-SELF) Trial participants (control, intervention and run-in participants) to explore ways to support mobile teledermatology uptake among underserved populations. Participants had 1+ underserved population characteristics: age &#x2265;70 years, residence outside major cities, low income, and Aboriginal and/or Torres Strait Islander (First Nations Australians) identity. We used thematic analysis to code data and report findings.ResultsOf the 22 participants (55% women; mean age 64.7 years), nine were &#x2265;70 years, 13 lived outside major cities, four had low income and two identified as First Nations Australians (six had &#x2265;2 characteristics). Participants valued the convenience of home-based skin checks, reassurance from expert feedback delivered via teledermatology and increased access to dermatology care. Family or partner support helped participants use the technology and examine hard-to-see areas, which increased confidence and adherence to the intervention. Barriers to skin self-examination included physical limitations, low confidence, and difficulties using the app or dermatoscope. Regional participants noted that teledermatology could potentially help reduce travel costs, provided additional clinic visits were not prompted by the teledermatology. Participants suggested that clear instructions and face-to-face training would facilitate uptake in practice, and that help from health professionals may also be beneficial.ConclusionTailored instructions, training, and technical assistance may help underserved groups use patient-led melanoma surveillance and improve their access to dermatological care.

This dataset provides information for 3235 Brachiopoda specimens held within the Ocean Benthic Fauna collection (collection code: OBFc) in the Shirshov Institute of Oceanology (IORAS). Collected from 1852 unique locations across the World Ocean, these specimens represent a 100-year record of Brachiopoda biodiversity, with sampling beginning in 1924.The IORAS Brachiopoda collection, due to its high species diversity, is a valuable resource for researchers. Broad taxonomic representation provides critical baseline data that fundamentally advance our understanding of brachiopod distribution patterns, ecological niches, and historical biogeography. The collection thus directly supports a wide range of disciplinary research, including taxonomy and phylogenetics, and the study of large-scale macroecological patterns in global marine ecosystems. A comprehensive revision and digitization of the complete Brachiopoda collection of the Shirshov Institute of Oceanology has yielded a detailed dataset on specimen distribution across geographic, bathymetric, and taxonomic categories. The project prioritized the documentation and imaging of the type specimens. This new dataset is a significant scientific resource for taxonomy, biodiversity, and biogeography; this data enhances our understanding of marine biodiversity and the distribution patterns of Brachiopoda throughout various oceanic regions and depth zones.

This study evaluates diagnostic coding consistency across sequential hospitalizations using a longitudinal analysis of the CMS Inpatient Dataset with more than five million records. By transforming ICD-10-CM codes into CCS categories, the research tracks 9 chronic and 7 acute conditions using State Transition Matrices to identify patterns: Continuous, Fading, Intermittent Gap, and Late Onset. Significant diagnostic gaps were observed. Diabetes and Multiple Myeloma showed 'Intermittent Gap' rates of approximately 14%. Critically, for most chronic conditions, the 'Intermittent Gap' pattern correlated with the highest hospital Length of Stay, suggesting that missed documentation may obscure patient complexity. While some gaps stem from coding regulations or clinical transitions (such as remission or combination coding) the prevalence of intermittent patterns indicates systemic coding failures, with possible clinical coordination implications.

The function and lifetime of moving mechanical assemblies (MMAs) in space depend on the properties of lubricants. MMAs that experience high speeds or high cycles require liquid-based lubricants due to their ability to reflow to the point of contact. However, only a few liquid-based lubricants have vapor pressures low enough for the vacuum conditions of space, each of which has limitations that add constraints to MMA designs. This work introduces a data-driven machine learning (ML) approach to predicting vapor pressure, enabling virtual screening and discovery of new space-suitable liquid lubricants. The ML models are trained with data from both high-throughput molecular dynamics simulations and experimental databases. The models are designed to prioritize interpretability, enabling the relationships between chemical structure and vapor pressure to be identified. Based on these insights, several candidate molecules are proposed that may have promise for future space lubricant applications in MMAs. SCIENTIFIC CONTRIBUTION: This work develops interpretable machine learning models for vapor pressure of low&#xa0;volatility molecules, identifying the structural features that drive volatility. The framework is accurate in the ultra-low-volatility regime where existing methods are unreliable, enabling virtual screening of candidate space lubricants. All code and data are publicly available, and the approach generalizes to materials discovery in other extreme environments.

The term 'housekeeping gene' evokes a sense of banality. These are the boring genes that hum in the background, impervious to cell state or stress. For a developmental biologist captivated by the impact of differentiation or environmental cues on gene expression, housekeeping genes find their use as loading controls. It is easy to lose sight of the fact that housekeeping genes are perhaps the most important genes - they code for the essential cellular factors: cytoskeleton components, basal metabolic enzymes, ribosomal proteins, etc. Therefore, understanding the basis of their extremely robust transcriptional regulation has critical implications for development and life in general.

Hospitals are high-risk environments where Indoor Air Quality (IAQ) significantly impacts the health and safety of personnel. Inefficient ventilation and the accumulation of pollutants can lead to Sick Building Syndrome and various non-communicable diseases. Conventional manual reporting systems often limit the speed and accuracy of risk assessment. Therefore, this study aimed to develop a digital solution to provide real-time air quality data for precise occupational risk management. This action research followed the Systems Development Life Cycle (SDLC) and Database Life Cycle (DBLC) through seven phases. An IoT-based monitoring device was developed to measure eight critical parameters: CO2, CO, TVOCs, PM2.5, PM10, temperature, and relative humidity. The system utilized Google Apps Script for automated reporting and Google App Sheet for data visualization, with data stored and processed through Google Sheets and a cloud server. The developed system successfully integrated IoT sensors with Google Workspace for real-time monitoring via mobile and web interfaces. Evaluation results indicated a high level of user satisfaction among hospital personnel, with mean scores ranging from 4.03 to 4.50. The highest ratings were achieved in "System Processing Speed" and "Alignment with User Requirements" (Mean = 4.50), demonstrating that the application effectively addresses staff needs. The application proves to be an efficient tool for monitoring hospital air quality and managing occupational health risks. By integrating IoT technology with low-code platforms, the system reduces administrative workload and promotes sustainable risk management practices.

To examine the relationship between metabolic syndrome (MetS) components and the development of malignant neoplasms of the brain (MNBs) and to identify the most significant influence. We conducted a nationwide cohort study using the Korean National Health Insurance Service database, enrolling 3, 976, 961 individuals aged &#x2265;40 years who completed a national health checkup in 2009. Participants were followed from 2010 to 2020 to assess MNB incidence using International Classification of Diseases, 10th Revision, codes C71.0-C71.9. Participants were stratified by the number of MetS components (0-5). Cox proportional hazards models were used to estimate hazard ratios (HRs), adjusted for age, sex, smoking, alcohol use, exercise, and comorbidities. In multivariate analysis adjusting for age, sex, lifestyle factors, and comorbidities, individual component counts modeled as discrete categorical comparisons against the zero-component reference did not reach statistical significance at any level (HR for 5 components: 1.070; 95% CI: 0.941-1.217; p = 0.303). However, when participants were stratified by clinically established cumulative thresholds, those with &#x2265;3 MetS components demonstrated a modestly but statistically significantly increased MNB risk (HR: 1.091; 95% CI: 1.017-1.170; &#x2265;5 components, 1.153; 95% CI: 1.040-1.278). Among individual components, elevated triglycerides showed the strongest independent association with MNB risk (adjusted HR: 1.517; 95% CI: 1.179-1.953), whereas the other components showed weaker or non-significant associations. The cumulative burden of MetS was modestly but statistically significantly associated with MNB risk, with elevated triglycerides showing the strongest association. These findings suggest that metabolic health management, particularly lipid monitoring, may inform brain tumor risk assessment. Critically, however, the outcome variable encompasses a diagnostically heterogeneous group of brain malignancies identified by administrative ICD-10 coding, and no mechanistic inference or histological subtype-specific conclusion can be drawn from the present data. These findings should be considered hypothesis-generating and require prospective validation using pathologically and molecularly characterized tumor registries before any clinical recommendations regarding lipid management for brain tumor prevention can be made.

Morbilliform drug eruptions in palliative care are often misattributed to active drugs, leading to inappropriate "opioid allergy" labels. Generic medications contain varying inactive excipients that can trigger delayed hypersensitivity. A man in his 70s with a prior history of ibuprofen-associated angioedema and metoclopramide-associated rash developed a pruritic morbilliform eruption after switching from a Mallinckrodt hydrocodone-acetaminophen product to a Camber oxycodone-acetaminophen formulation (National Drug Code [NDC] 31722-0951-05). During a washout period, the patient successfully tolerated SpecGx-manufactured morphine (NDC 00406-5118-01) and Aurobindo potassium chloride (NDC 65862-0987-99) while the rash resolved. Comparison of NDC data identified croscarmellose sodium as the likely culprit, as it was unique to the offending formulation. Clinicians should document the NDC rather than broad drug classes when reactions occur after a generic substitution to preserve future therapeutic options and ensure continued access to essential analgesia.

The objective of this study is to evaluate the ethical dimensions of reflex and reflective testing (RRT) within the framework of value-based laboratory medicine and in alignment with ISO 15189:2022. The study aims to highlight the challenges in guiding and standardizing these practices in an era of value-based medicine. The ISO 15189:2022 requirements for quality and competence were examined alongside the IFCC Code of Ethics and relevant literature. The ethical evaluation was structured around the four core principles of bioethics: beneficence, non-maleficence, autonomy, and justice. Specific ethical issues related to RRT were identified through expert discussion and compared with the ISO standard. The thorough examination reveals that although ISO 15189:2022 does not use the RRT terminology directly, its requirements provide a robust framework for these interventions. Based on beneficence, non-maleficence, autonomy, and justice, RRT interventions align with ISO 15189:2022 and promote value-based laboratory medicine. Laboratory specialists/professionals must balance the four ethical principles to ensure that the "total testing process" serves the patient's best interest in conformity with ISO 15189:2022. By adopting a strong ethical culture, the profession can successfully transition from being a producer of test results to a key partner in personalized and value-based medicine.

Early and accurate diagnosis of Alzheimer's disease (AD) is critical. In MRI-based computer-aided diagnosis, convolutional neural networks (CNNs) excel at extracting local features but struggle to model long-range dependencies, while Vision Transformers (ViTs) offer strong global modeling capabilities but suffer from high computational complexity, limiting their deployment in resource-constrained settings. This paper proposes HDFT-MViT, a lightweight hybrid architecture based on MobileViT that integrates a hierarchical dynamic filter with a lightweight Transformer. The model adopts a progressive Core-Enhanced Mix design: Shallow layers employ MobileNetV2 inverted residual blocks for efficient local feature extraction; intermediate and deep layers incorporate a dual-branch module that integrates a dynamic filter for frequency-domain global modulation and a lightweight Transformer for spatial long-range dependency modeling, followed by hierarchical fusion via learnable weights. A channel attention mechanism is further introduced to enhance feature discriminability. Evaluations on the public ADNI-1 (3-class) and ADNI-2 (4-class) MRI datasets show that HDFT-MViT achieves state-of-the-art classification accuracies of 98.85&#x202f;&#xb1;&#x202f;0.27% and 98.07&#x202f;&#xb1;&#x202f;0.54%, respectively, while maintaining a lightweight profile with only 3.46 M parameters, confirming its effectiveness and efficiency. HDFT-MViT achieves an optimal balance between local detail perception and global semantic understanding within a computationally efficient framework, offering a promising tool for clinical AD diagnosis. Code will be released upon acceptance.