Identity details within an ECG is jointly situated within local and global features. The current methods for ECG recognition emphasize only on local or global details. They have also paid limited attention to unseen and cross-domain scenarios. Furthermore, there exists a lack of consensus on evaluation strategies. Thus, this paper introduces LGTraNet, a generalized architecture designed to establish baselines for securing personal identity using ECG biometrics in cross-domain scenarios. Our proposed model firstly extracts identity details at local temporal levels. The extracted features are then calibrated with globally details using a Self-Calibrated Normalizing Residual Network (SCNRNet). Finally, the refined local details are aggregated using a transformer model to formulate robust global identity representations. We evaluate LGTraNet over challenging cross-domain scenarios, such as cross-session and cross-database. To mitigate challenges in domain-shift, we also introduce an transfer learning based training strategy. Experimental study conducted on three benchmark datasets, ECG1D, MIT-BIH, and PTB, shows that the LGTraNet achieves significant performance in cross-domain settings, and outperforms state-of-the-art. Our code is available at: https://github.com/AmanVerma2307/LGTraNet.

We present the delta-gesture biometrics quantification assessment (DGBQA) framework which estimates the biometric characteristics of hand gestures. The proposed framework is aimed at learning generic motion-representations of gestures instead of subject-specific details from a large number of identities. It also enables the biometric scores to be estimated for a set of gestures at a time instead of having to estimate these one at a time. In the first step, it formulates a feature space which is identity and gesture aware, and in the second step, it proceeds to compute biometric scores using inter-subject and intra-subject distance measures in the feature space. However, due to the inclusion of identity-aware objective, the identity details tend to be shared across gestures. We refer to this as identity sharing and this can lead to the score for different gestures being dependent on each other. To address this issue, we introduce an identity-level cross-gesture disentanglement loss$(\mathscr{L}_{ICGD})$which encourages the different gestures belonging to the same identity to be orthogonal in the feature space. We demonstrate the efficacy of the proposed biometric quantification framework and the disentanglement loss function through extensive experiments on four datasets and using standard as well as proposed novel evaluation metrics. Our analysis indicates that gestures involving multiple coarse movements are better for biometrics.

With almost half the Indian population facing water stress [ 42 ], likely to be aggravated with climate change [ 76 , 122 ], an intersecting crises of environment, livelihood, and social justice can worsen the vulnerability of rural communities that are directly dependent upon water resources for their livelihood and sustenance. To address this challenge of rural water security especially for marginalized communities, MGNREGA is a public works and employment generation scheme in India that contributes towards the construction of water related assets such as farm ponds, checkdams, and trenches and bunds, to improve irrigation and climate resilience of rural communities. MGNREGA is a demand-driven scheme that solicits demands for water related assets through the local governance structure in India of village-level community meetings called Gram Sabhas. In many areas, Civil Society Organizations (CSOs) and community based institutions also play a significant role by coordinating participatory processes to build the capacity of rural community members to participate in local governance and demand such assets under MGNREGA. However, through multiple field visits, we observed a need to strengthen local processes with appropriate data and tools to support context-sensitive and evidence-based bottom-up planning. In this study, we adopt an approach of building a digital participatory tool, built on a technology stack that encapsulates different geospatial datasets, to answer key research questions of whether such technology artefacts can improve the participation of rural communities in water security planning and management, through, (1) community empowerment by building a data-driven understanding of water security, (2) ecological sustainability of water resources through scientific planning of assets, and (3) equity led participatory processes to improve fairness in the distribution of assets. We discuss here the design of the tool, Commons Connect, underlying datasets on which it is built, and findings from an initial field testing of the tool across four locations in India. We outline several improvements and open questions that remain, and highlight how such tools can be integrated into the MGNREGA operational workflow and improve the sustainability and resilience of rural social-ecological systems.

Electronic health records (EHRs) provide rich clinical information and the opportunities to extract epidemiological patterns to understand and predict patient disease risks with suitable machine learning methods such as topic models. However, existing topic models do not generate identifiable topics each predicting a unique phenotype. One promising direction is to use known phenotype concepts to guide topic inference. We present a seed-guided Bayesian topic model called MixEHR-Seed with 3 contributions: (1) for each phenotype, we infer a dual-form of topic distribution: a seed-topic distribution over a small set of key EHR codes and a regular topic distribution over the entire EHR vocabulary; (2) we model age-dependent disease progression as Markovian dynamic topic priors; (3) we infer seed-guided multi-modal topics over distinct EHR data types. For inference, we developed a variational inference algorithm. Using MixEHR-Seed, we inferred 1569 PheCode-guided phenotype topics from an EHR database in Quebec, Canada covering 1.3 million patients for up to 20-year follow-up with 122 million records for 8539 and 1126 unique diagnostic and drug codes, respectively. We observed (1) accurate phenotype prediction by the guided topics, (2) clinically relevant PheCode-guided disease topics, (3) meaningful age-dependent disease prevalence. Source code is available at GitHub: https://github.com/li-lab-mcgill/MixEHR-Seed.

Electronic Health Records (EHRs) contain rich clinical data collected at the point of the care, and their increasing adoption offers exciting opportunities for clinical informatics, disease risk prediction, and personalized treatment recommendation. However, effective use of EHR data for research and clinical decision support is often hampered by a lack of reliable disease labels. To compile gold-standard labels, researchers often rely on clinical experts to develop rule-based phenotyping algorithms from billing codes and other surrogate features. This process is tedious and error-prone due to recall and observer biases in how codes and measures are selected, and some phenotypes are incompletely captured by a handful of surrogate features. To address this challenge, we present a novel automatic phenotyping model called MixEHR-Guided (MixEHR-G), a multimodal hierarchical Bayesian topic model that efficiently models the EHR generative process by identifying latent phenotype structure in the data. Unlike existing topic modeling algorithms wherein the inferred topics are not identifiable, MixEHR-G uses prior information from informative surrogate features to align topics with known phenotypes. We applied MixEHR-G to an openly-available EHR dataset of 38,597 intensive care patients (MIMIC-III) in Boston, USA and to administrative claims data for a population-based cohort (PopHR) of 1.3 million people in Quebec, Canada. Qualitatively, we demonstrate that MixEHR-G learns interpretable phenotypes and yields meaningful insights about phenotype similarities, comorbidities, and epidemiological associations. Quantitatively, MixEHR-G outperforms existing unsupervised phenotyping methods on a phenotype label annotation task, and it can accurately estimate relative phenotype prevalence functions without gold-standard phenotype information. Altogether, MixEHR-G is an important step towards building an interpretable and automated phenotyping system using EHR data.

BACKGROUND: Venous thromboembolisms (VTEs), which include deep vein thrombosis (DVT) and pulmonary embolism (PE), are associated with significant mortality, morbidity, and cost in hospitalized patients. To evaluate the success of preventive measures, accurate and efficient methods for monitoring VTE rates are needed. Therefore, we sought to determine the accuracy of statistical natural language processing (NLP) for identifying DVT and PE from electronic health record data. METHODS: We randomly sampled 2000 narrative radiology reports from patients with a suspected DVT/PE in Montreal (Canada) between 2008 and 2012. We manually identified DVT/PE within each report, which served as our reference standard. Using a bag-of-words approach, we trained 10 alternative support vector machine (SVM) models predicting DVT, and 10 predicting PE. SVM training and testing was performed with nested 10-fold cross-validation, and the average accuracy of each model was measured and compared. RESULTS: On manual review, 324 (16.2%) reports were DVT-positive and 154 (7.7%) were PE-positive. The best DVT model achieved an average sensitivity of 0.80 (95% CI 0.76 to 0.85), specificity of 0.98 (98% CI 0.97 to 0.99), positive predictive value (PPV) of 0.89 (95% CI 0.85 to 0.93), and an area under the curve (AUC) of 0.98 (95% CI 0.97 to 0.99). The best PE model achieved sensitivity of 0.79 (95% CI 0.73 to 0.85), specificity of 0.99 (95% CI 0.98 to 0.99), PPV of 0.84 (95% CI 0.75 to 0.92), and AUC of 0.99 (95% CI 0.98 to 1.00). CONCLUSIONS: Statistical NLP can accurately identify VTE from narrative radiology reports.

Unplanned hospital readmissions raise health care costs and cause significant distress to patients. Hence, predicting which patients are at risk to be readmitted is of great interest. In this paper, we mine large amounts of administrative information from claim data, including patients demographics, dispensed drugs, medical or surgical procedures performed, and medical diagnosis, in order to predict readmission using supervised learning methods. Our objective is to gain knowledge about the predictive power of the available information. Our preliminary results on data from the provincial hospital system in Quebec illustrate the potential for this approach to reveal important information on factors that trigger hospital readmission. Our findings suggest that a substantial portion of readmissions is inherently hard to predict. Consequently, the use of the raw readmission rate as an indicator of the quality of provided care might not be appropriate.