Smart Health最新文献

{"title":"Multi-level cancer profiling through joint cell-graph representations","authors":"Luis Carlos Rivera Monroy ,&nbsp;Leonhard Rist ,&nbsp;Frauke Wilm ,&nbsp;Christian Ostalecki ,&nbsp;Andreas Baur ,&nbsp;Julio Vera ,&nbsp;Katharina Breininger ,&nbsp;Andreas Maier","doi":"10.1016/j.smhl.2024.100470","DOIUrl":"https://doi.org/10.1016/j.smhl.2024.100470","url":null,"abstract":"<div><p>Computer-aided analysis of digitized pathology samples has significantly advanced with the rapid progression of machine and Deep Learning (DL) methods. However, most existing approaches primarily focus on features extracted from patches due to the large image sizes. This focus limits the ability of Convolutional Neural Networks (CNNs) to capture global information from the samples, resulting in an incomplete phenotypical and topological representation and thereby restricting the diagnostic capabilities of these methods. The recent emergence of Graph Neural Networks (GNNs) offers new opportunities to overcome these limitations through graph-driven representations of pathological samples. This work introduces a graph-based framework that encompasses diverse cancer types and integrates different imaging modalities. In this framework, histopathology samples are represented as graphs, and a pipeline facilitating cell-wise and disease classification is developed. The results support this motivation: for cell-wise classification, we achieved an average accuracy of <span><math><mrow><mn>88</mn><mtext>%</mtext></mrow></math></span>, and for disease-wise classification, an average accuracy of 83%, outperforming reference models such as XGBoost and standard CNNs. This approach not only provides flexibility in combining various diseases but also extends to integrating different staining techniques.</p></div>","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"32 ","pages":"Article 100470"},"PeriodicalIF":0.0,"publicationDate":"2024-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352648324000266/pdfft?md5=7c470921f13bec160c26b55acac4f145&pid=1-s2.0-S2352648324000266-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140290856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Adaptive attention-aware fusion for human-in-the-loop behavioral health detection","authors":"Martin Brown ,&nbsp;Abm Adnan Azmee ,&nbsp;Md. Abdullah Al Hafiz Khan ,&nbsp;Dominic Thomas ,&nbsp;Yong Pei ,&nbsp;Monica Nandan","doi":"10.1016/j.smhl.2024.100475","DOIUrl":"https://doi.org/10.1016/j.smhl.2024.100475","url":null,"abstract":"<div><p>Identifying behavioral health is paramount for law enforcement officers to provide appropriate follow-up community care. In the current practice, law enforcement offices manually identify these behavioral health cases to allow the designation of the relevant follow-up resources. In this work, we develop a tool to automatically detect behavioral health cases from police public narrative reports by identifying behavioral health indicator signals. We propose a novel adaptive attention-aware fusion model for detecting behavioral health signals in sensitive police reports. Our model leverages contextual and semantic information from the reports and relevant behavioral health cues as keywords from a pre-trained attention-weighted keyword-based model. Our model also employs label self-attention mechanisms to correlate label embeddings with the report and keyword representations. Furthermore, we propose a novel clustering-based uncertainty-enabled informative sampling query strategy to integrate humans-in-the-loop in the active learning framework to reduce required annotation from experts. This querying strategy selects the most informative and diverse samples for expert annotation. Our experimental results showed that the proposed model outperforms state-of-the-art classifiers on a dataset of 300 manually annotated ground truth police reports, achieving an accuracy of 87.58% and an F1-score of 85.67%. Applying our querying strategy to our proposed model increased the detection of behavioral health, achieving an accuracy of 92% and an F1-score of 91.1%. Also, our proposed model achieves an accuracy score of 93.75% and an F1-score of 93.61% on unseen samples. Lastly, our proposed model demonstrates its interpretability by extracting the keywords associated with each behavioral health category.</p></div>","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"32 ","pages":"Article 100475"},"PeriodicalIF":0.0,"publicationDate":"2024-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140343877","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Explainable AI and transformer models: Unraveling the nutritional influences on Alzheimer's disease mortality","authors":"Ziming Liu ,&nbsp;Longjian Liu ,&nbsp;Robert E. Heidel ,&nbsp;Xiaopeng Zhao","doi":"10.1016/j.smhl.2024.100478","DOIUrl":"https://doi.org/10.1016/j.smhl.2024.100478","url":null,"abstract":"<div><p>This pioneering study introduces the use of transformer-based machine learning models and explainable AI approaches to explore the impact of nutrition on Alzheimer's disease (AD) mortality. Using data from the Third National Health and Nutrition Examination Survey (Nhanes iii 1988 to 1994) and the NHANES III Mortality-Linked File (2019) databases, we investigate the intricate relationship between various nutritional factors and AD mortality. Our approach features a novel application of transformer models, which are then benchmarked against established methods like random forests and support vector machines. This comparison not only underscores the strengths of transformer models in handling complex medical datasets but also highlights their potential for providing deeper insights into disease progression. Key findings, such as the significant roles of Platelet distribution width in AD mortality in transformer and Serum Vitamin B12 in random forest, are enhanced by the use of Explainable Artificial Intelligence (XAI), particularly the Shapley Additive Explanations (SHAP) and the integrated gradient methods. This study serves as a vital step forward in applying advanced AI techniques to medical research, offering new perspectives in understanding and combating Alzheimer's Disease.</p></div>","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"32 ","pages":"Article 100478"},"PeriodicalIF":0.0,"publicationDate":"2024-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140320797","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Application of time series analysis to classify therapeutic breathing patterns","authors":"João Lucas Oliveira Canhoto ,&nbsp;Paulo Salgado Gomes de Mattos Neto ,&nbsp;Taiwan Roberto Barbosa ,&nbsp;José Emmanuel Matias da Silva Santos ,&nbsp;Igor Mauricio de Campos ,&nbsp;Geraldo Leite Maia Junior ,&nbsp;João Victor Cordeiro Coutinho ,&nbsp;Márcio Evaristo da Cruz Brito ,&nbsp;Anna Luisa Araújo Brito ,&nbsp;Daniella Cunha Brandão ,&nbsp;Armele de Fátima Dornelas de Andrade ,&nbsp;Herbert Albérico de Sá Leitão ,&nbsp;Shirley Lima Campos","doi":"10.1016/j.smhl.2024.100460","DOIUrl":"10.1016/j.smhl.2024.100460","url":null,"abstract":"<div><h3>Objective</h3><p>Compare various methods for measuring time series similarity in order to classify referenced therapeutic breathing patterns (BP) used in respiratory disorder rehabilitation.</p></div><div><h3>Methods</h3><p>This experimental study involved the collection of respiratory signals during specified breathing exercises conducted with healthy volunteers. The study employed a screening phase using a k-NN classifier and eight distance measurement methods, including Minkowski Distance, Dynamic Time Warping-DTW (including FastDTW and constrained-cDTW variations), Longest Common Subsequence-LCSS, Edit Distance on Real Sequences-EDR, Time Warp Edit Distance-TWEED, and Minimum Jump Costs-MJC. Two distinct approaches were employed for classifying therapeutic BP based on time series similarity: (1) using the k-Shape algorithm for clustering, and 2) integrating methods to represent therapeutic BP and classify test curves using the most relevant measurement methods obtained from the first approach.</p></div><div><h3>Results</h3><p>Among the two tested approaches, the combination of the cDTW algorithm and Minkowski distance (p = 2), using the 1-NN classifier, achieved the highest scores in this study, closely matching the metrics obtained from visual inspection conducted by human evaluators.</p></div><div><h3>Conclusion</h3><p>The use of combined classification methods in the analysis of flow curves referring to therapeutic breathing patterns improves the classification results, with metrics closely aligned with those obtained through visual evaluation conducted by individuals.</p></div><div><h3>Significance</h3><p>Time series analysis methods proved to be sensitive to classify respiratory flow curves equivalent to therapeutic breathing patterns used in respiratory disorder rehabilitation. This methodology can be used to monitor respiratory curves in new applications and implementation in devices for evaluating and treating the ventilatory pattern.</p></div>","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"32 ","pages":"Article 100460"},"PeriodicalIF":0.0,"publicationDate":"2024-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140091152","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Few-shot meta-learning for pre-symptomatic detection of Covid-19 from limited health tracker data","authors":"Atifa Sarwar,&nbsp;Abdulsalam Almadani,&nbsp;Emmanuel O. Agu","doi":"10.1016/j.smhl.2024.100459","DOIUrl":"https://doi.org/10.1016/j.smhl.2024.100459","url":null,"abstract":"<div><p>Detecting (or screening for) Covid-19 even before symptoms fully manifest, could enable patients to receive timely and life-saving treatment. Prior work has demonstrated that heart rate and step data from low-end wearables analyzed using deep learning models can detect Covid-19 reliably. However, significant individual differences in vital sign manifestation (high inter-subject variability) present a challenge to the generalization of deep learning models across diverse users. The limited amount of data in many medical scenarios further exacerbates this issue. Consequently, neural network models that can learn from limited vital sign data and varied inter-subject patterns are compelling. Meta-learning has emerged as a powerful technique for tackling various machine learning challenges, including insufficient data, domain shifts across datasets, and issues with generalization. This study proposes <em>MetaCovid</em>, a deep adaptation framework that employs meta-learning to address the variability of vital sign manifestation between subjects using only two days of data in order to detect Covid-19 before symptoms manifest. <em>MetaCovid</em> leverages heart rate and step measurements collected from consumer-grade health trackers over the preceding 2 days, extracts 45 digital bio-markers (features), which along with raw data, are fed into a deep GRU-based network with an attention mechanism, followed by uncertainty filtering. <em>MetaCovid</em> is trained using OC-MAML, a one-class few-shot MAML variant that adapts to the target distribution/user using only samples from the majority class. <em>MetaCovid</em> generalized well across two relatively small, publicly available Covid-19 datasets, achieving a recall of 0.81 and 0.92, and detecting 61% (14 out of 23) and 50% (17 out of 34) of users infected with Covid-19 before symptom onset. When OC-MAML was excluded from <em>MetaCovid</em> in an ablation study, the F₂ score dropped by 36%, highlighting that meta-learning indeed facilitates adaptation of deep sensing models to varying vital sign patterns. Notably, <em>MetaCovid</em> outperforms the current state-of-art method by predicting Covid-19 early on day <span><math><mi>N</mi></math></span> using heart rate and step measurements from only the preceding 2 days compared to 28 days, reducing data requirements by 93%. To the best of our knowledge, our study is the first to propose utilizing meta-learning to mitigate vital sign variability with limited data for Covid-19 screening. We believe that <em>MetaCovid</em> will pave the way for innovative Covid-19 interventions that are accurate even with limited data and help contain the spread of infectious diseases in the future.</p></div>","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"32 ","pages":"Article 100459"},"PeriodicalIF":0.0,"publicationDate":"2024-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140031506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Retraction notice to ‘Blockchain technology in the future of healthcare’ [Smart Health 23 (2022) 100223]","authors":"Quazi Mamun","doi":"10.1016/j.smhl.2024.100458","DOIUrl":"https://doi.org/10.1016/j.smhl.2024.100458","url":null,"abstract":"","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"32 ","pages":"Article 100458"},"PeriodicalIF":0.0,"publicationDate":"2024-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S235264832400014X/pdfft?md5=fac559044334cbd8c0b8378ca18a3a66&pid=1-s2.0-S235264832400014X-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139999226","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Recent trends and techniques of blood glucose level prediction for diabetes control","authors":"Benzir Md. Ahmed ,&nbsp;Mohammed Eunus Ali ,&nbsp;Mohammad Mehedy Masud ,&nbsp;Mahmuda Naznin","doi":"10.1016/j.smhl.2024.100457","DOIUrl":"10.1016/j.smhl.2024.100457","url":null,"abstract":"<div><p>Diabetes, a metabolic disorder disease, can cause short-term acute or even long-term chronic complications in a patient’s body. In 2021, 10.5% of the world’s adult population had diabetes. These numbers are increasing day by day, which results in an associated increase of morbidity, mortality, and health care cost related to diabetes. Thus, a huge research effort has been carried out to manage diabetes. A precursor to diabetes management is to predict the future blood glucose levels based on a patient’s past history. In this paper, we provide a comprehensive and systematic study of diabetes management, focusing on recent research towards blood glucose level prediction. In particular, we have categorized and presented existing recent research based on major clinical application domains, different input features, and major modeling techniques including physiological, data-driven, and hybrid models. We have summarized the performance analysis of different modeling techniques using different metrics, and critically analyzed these techniques from different perspectives. Finally, we have identified a number of research challenges and potential future works that range from data collection to model improvement for Type 2 Diabetes Mellitus. This review can be a good starting point for researchers and practitioners who are working in building data-driven computational models for diabetes management and blood glucose level prediction.</p></div>","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"32 ","pages":"Article 100457"},"PeriodicalIF":0.0,"publicationDate":"2024-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139880929","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Recent trends and techniques of blood glucose level prediction for diabetes control","authors":"B. M. Ahmed, Mohammed Eunus Ali, Mohammad Mehedy Masud, Mahmuda Naznin","doi":"10.1016/j.smhl.2024.100457","DOIUrl":"https://doi.org/10.1016/j.smhl.2024.100457","url":null,"abstract":"","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"551 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139820696","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Design and technical evaluation of an AMBU-BAG based low-cost ventilator-AARMED","authors":"Mohit Kumar ,&nbsp;Ravinder Kumar ,&nbsp;Vishal Kumar ,&nbsp;Amanpreet Chander ,&nbsp;Abhinav Airan ,&nbsp;Rajesh Arya ,&nbsp;Gurpreet Singh Wander ,&nbsp;Ashish Kumar Sahani","doi":"10.1016/j.smhl.2023.100445","DOIUrl":"https://doi.org/10.1016/j.smhl.2023.100445","url":null,"abstract":"<div><p>The COVID-19 pandemic has caused a significant strain on the healthcare system worldwide, resulting in an acute shortage of ventilators. Conventional ventilators are costly, and production is difficult to scale up during a rapidly spreading pandemic. Ambu bags offer a low-cost solution for manual ventilation, but their lack of precise control over parameters makes them unsuitable as a replacement for conventional ventilators. To address this issue, we propose the AARMED (Ambu bag Attachment for Rapid Mass Emergency Deployment) system, which is a low-cost and easy-to-assemble mechanical resuscitator that can achieve most of the recommended modes and parameter ranges for managing COVID-19 patients. AARMED can operate in volume-control, pressure-control, and assist-controlled modes continuously over long periods, making it suitable for use in emergency settings. The AARMED system has been tested using an ISO certified Test-lung over various parameter settings and has been found to be an effective alternative to costly conventional ventilators. It has a battery backup of 2.5 h under normal operating conditions, making it an ideal transport ventilator. In conclusion, the AARMED system offers a low-cost and easy-to-assemble solution for mechanical ventilation in emergency settings. Its ability to achieve most of the recommended modes and parameter ranges for managing COVID-19 patients makes it a viable alternative to conventional ventilators in resource-constrained settings.</p></div>","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"31 ","pages":"Article 100445"},"PeriodicalIF":0.0,"publicationDate":"2024-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352648323000739/pdfft?md5=687d1e050cacbb7658c4f24834b63b29&pid=1-s2.0-S2352648323000739-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139504670","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CardioHelp: A smartphone application for beat-by-beat ECG signal analysis for real-time cardiac disease detection using edge-computing AI classifiers","authors":"Ucchwas Talukder Utsha,&nbsp;Bashir I. Morshed","doi":"10.1016/j.smhl.2024.100446","DOIUrl":"https://doi.org/10.1016/j.smhl.2024.100446","url":null,"abstract":"<div><p>Cardiovascular diseases are a leading cause of morbidity and mortality worldwide. To diagnose cardiac diseases, physicians often utilize a combination of medical history, physical examination, and several diagnostic tests, such as electrocardiograms (ECG/EKG), echocardiograms, and stress tests. Early detection and effective management of cardiac diseases play a crucial role in improving patient outcomes and reducing healthcare burden. To address this concern, we introduce a novel edge-computing approach for cardiac healthcare using a smartphone application (CardioHelp) that combines heart rate monitoring with the detection of abnormal heartbeats in individuals. Our approach centers around a user-friendly smart-health application designed to visualize ECG signals, track and monitor heart rate continuously, and recognize and notify users of any anomalies through advanced beat-by-beat ECG analysis algorithms and artificial intelligence (AI) techniques including machine learning and deep learning. Our system includes a custom wearable ECG data collection system that can transfer data to CardioHelp in real-time. In this study, we have used the MIT-BIH Arrhythmia dataset to train deep learning models using intricate patterns and features representative of various heart conditions. Among the deep learning models, the Long Short-Term Memory (LSTM) demonstrated superior performance, obtaining an accuracy of 98.74% and precision and recall of 99.95% and 99.86%, respectively. By transferring the MIT-BIH Arrhythmia Database’s test dataset through our application as mock real-time data, we assessed our CardioHelp application’s accuracy in identifying and classifying various heart conditions. The LSTM model is found to be the most accurate model providing an accuracy of 95.94% for ECG beat classification. The results confirmed the effectiveness of our developed smartphone system, demonstrating its ability to accurately detect and classify cardiac conditions. As our novel approach presents a complimentary cardiac healthcare system using a smart health application, this CardioHelp has the potential to significantly enhance preventive care, enable early intervention, and improve overall cardiovascular health outcomes.</p></div>","PeriodicalId":37151,"journal":{"name":"Smart Health","volume":"31 ","pages":"Article 100446"},"PeriodicalIF":0.0,"publicationDate":"2024-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352648324000011/pdfft?md5=9c56788fa103efe37a12332fec7595c8&pid=1-s2.0-S2352648324000011-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139493933","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}