Smart Health最新文献

The extent of neurophysiological activation, such as brain activities, eye movement, and skin conductance, can vary as a joint function of cognitive load. These functions are the basis of models that describe human behavior and neural mechanisms for diagnosing and treating cognitive disorders, such as Alzheimer’s disease, mild cognitive impairment, and stroke-related cognitive dysfunction. Such models can enhance our understanding of the disease processes and enable crucial applications like predicting cognitive trajectories for early diagnosis. However, despite the success of these models in predicting early-stage cognitive impairment and decline, their practical use is limited in clinics because most of the models focus on utilizing one or two factors of the neurophysiological activation to achieve prediction. Still, little is known about the mechanisms through which the task difficulty and cognitive demands affect the expression of neurophysiological activation and whether there is an expression difference under cognitive task demands. The purpose of this paper is to provide a comprehensive examination of the neurophysiological expression difference and mechanisms under various cognitive loads. We designed an experimental protocol and developed a data processing framework to explicitly examine brain activity and eye movement under various levels of cognitive task difficulty and find that (1) eye movement is a readout of cognitive processes, but it is a joint function of task difficulty, brain activity, and skin conductance; (2) brain activity has specific patterns related to the various levels of cognitive load and exerts its influence on predicting the dynamics of cognitive processes. These findings suggest that neuroimaging studies comparing task-related neurophysiological activation must be examined and interpreted in a holistic view of neural mechanisms.

Healthcare 4.0 is one of the emerging concepts that has grabbed the interest among researchers as well as the medical sector. Using the Internet of Things (IoT) and sophisticated communication technologies, it is now possible to monitor the patient from a remote area. In this paper, we design a remote health monitoring system using IoT and Machine Learning (ML) to determine the health condition of a patient. Supervised ML algorithms along with a time-series model such as Seasonal Autoregressive Integrated Moving Average (SARIMA) model are applied on the gathered data from IoT medical sensors to predict the health status of a patient. We consider a use-case of covid and compared it with our sensor data by applying the unsupervised ML algorithm, Long Short Term Memory (LSTM) along with a stochastic model, namely Markov Model to detect the risk of getting covid for a particular patient. LSTM with Markov model provides better results for detection with root mean squared error (RMSE) of 0.18 as against the RMSE of 0.45 obtained with only LSTM. We further design an optimization algorithm using “fuzzy logic” that attains optimum results in detecting the risk of getting covid.

A successful pregnancy is a contingent upon a complex network of interdependent biological adaptations, including maternal immune responses and hormonal balance. Recent improvements in high-intelligence technology have enabled the combination of clinical and social data with multi-omics biological data can offer the opportunity to detect maternal risk during pregnancy. The United Nations' sustainable development goal (SDG 3) aims to improve maternal health and reduce child and maternal mortality by 2030. Nevertheless, maternal mortality has not decreased at the indicated rate, especially in developing countries like Ghana. This paper aims to establish an intelligent machine learning-based system for effectively monitoring and predicting pregnant women's risk levels. We assessed pregnant women's health data and risk variables to determine the maternal risk intensity level during pregnancy. Therefore, we proposed a hybrid ensemble algorithm (XGBoost and CatBoost) to determine the significant health factors associated with maternal health and predict the mother's risk level during pregnancy. The study outcome showed that blood sugar, age and body temperature were the most significant factors in determining the risk level of a pregnant woman in MM. Also, the prediction outcome (accuracy of 93.99%, AUC of 96.96%, recall 92.44%, and precision 93.46%) shows that our model performed well compared with other studies and machine learning algorithms.

Serology and molecular tests are the two most commonly used methods for rapid COVID-19 infection testing. The two types of tests have different mechanisms to detect infection, by measuring the presence of viral SARS-CoV-2 RNA (molecular test) or detecting the presence of antibodies triggered by the SARS-CoV-2 virus (serology test). A handful of studies have shown that symptoms, combined with demographic and/or diagnosis features, can be helpful for the prediction of COVID-19 test outcomes. However, due to nature of the test, serology and molecular tests vary significantly. There is no existing study on the correlation between serology and molecular tests, and what type of symptoms are the key factors indicating the COVID-19 positive tests.

In this study, we propose a machine learning based approach to study serology and molecular tests, and use features to predict test outcomes. A total of 2,467 donors, each tested using one or multiple types of COVID-19 tests, are collected as our testbed. By cross checking test types and results, we study correlation between serology and molecular tests. For test outcome prediction, we label 2,467 donors as positive or negative, by using their serology or molecular test results, and create symptom features to represent each donor for learning. Because COVID-19 produces a wide range of symptoms and the data collection process is essentially error prone, we group similar symptoms into bins. This decreases the feature space and sparsity. Using binned symptoms, combined with demographic features, we train five classification algorithms to predict COVID-19 test results. Experiments show that XGBoost achieves the best performance with 76.85% accuracy and 81.4% AUC scores, demonstrating that symptoms are indeed helpful for predicting COVID-19 test outcomes. Our study investigates the relationship between serology and molecular tests, identifies meaningful symptom features associated with COVID-19 infection, and also provides a way for rapid screening and cost effective detection of COVID-19 infection.

The accurate segmentation of computed tomography (CT) scan volume is an essential step in radiomic analysis as well as in developing advanced surgical planning techniques with numerous medical applications. When this process is performed manually by a clinician, it is laborious, time consuming, prone to error, and its success depends to a large extent on the level of experience. In this work, we propose an automated deep learning (DL) segmentation framework for CT images called U-Net CT Segmentation (U-NetCTS) to combine the DL U-Net and CT images in the domain of automatic segmentation. Experimental results show that U-NetCTS framework can segment different CT DICOM image regions of interest in a range of random CT volumes. A statistical and qualitative comparison of the CT slices automatically segmented by U-NetCTS framework and ground-truth images indicates that U-NetCTS framework achieves a high level of accuracy, where the Tanimoto coefficient, dice similarity coefficient, and peak signal-to-noise ratio values are 99.06%, 99.52%, and 53.29 dB, respectively. The DC value is also higher than that of state-of-the-art DL techniques for automatic segmentation of CT images of various human organs. Furthermore, a total amount of 3595 CT slices is employed in this study with various CT region of interest to validate the results.

Acoustic signals generated by the human body have often been used as biomarkers to diagnose and monitor diseases. As the pathogenesis of COVID-19 indicates impairments in the respiratory system, digital acoustic biomarkers of COVID-19 are under investigation. In this paper, we explore an accurate and explainable COVID-19 diagnosis approach based on human speech, cough, and breath data using the power of machine learning. We first analyze our design space considerations from the data aspect and model aspect. Then, we perform data augmentation, Mel-spectrogram transformation, and develop a deep residual architecture-based model for prediction. Experimental results show that our system outperforms the baseline, with the ROC-AUC result increased by 5.47%. Finally, we perform an interpretation analysis based on the visualization of the activation map to further validate the model.

Background

Medication nonadherence is a critical problem with severe implications in indi-viduals at risk for atherosclerotic cardiovascular disease. Many studies have attempted to predict medication adherence in this population, but few, if any, have been effective in prediction, sug-gesting that essential risk factors remain unidentified.

Objective

This study's objective was to (1) establish an accurate prediction model of medi-cation adherence in individuals at risk for atherosclerotic cardiovascular disease and (2) identify significant contributing factors to the predictive accuracy of medication adherence. In particular, we aimed to use only the baseline questionnaire data to assess medication adherence prediction feasibility.

Methods

A sample of 40 individuals at risk for atherosclerotic cardiovascular disease was recruited for an eight-week feasibility study. After collecting baseline data, we recorded data from a pillbox that sent events to a cloud-based server. Health measures and medication use events were analyzed using machine learning algorithms to identify variables that best predict medication adherence.

Results

Our adherence prediction model, based on only the ten most relevant variables, achieved an average error rate of 12.9%. Medication adherence was closely correlated with being encouraged to play an active role in their treatment, having confidence about what to do in an emergency, knowledge about their medications, and having a special person in their life.

Conclusions

Our results showed the significance of clinical and psychosocial factors for pre-dicting medication adherence in people at risk for atherosclerotic cardiovascular diseases. Clini-cians and researchers can use these factors to stratify individuals to make evidence-based decisions to reduce the risks.

The large amount of data generated during the COVID-19 pandemic requires advanced tools for the long-term prediction of risk factors associated with COVID-19 mortality with higher accuracy. Machine learning (ML) methods directly address this topic and are essential tools to guide public health interventions. Here, we used ML to investigate the importance of demographic and clinical variables on COVID-19 mortality. We also analyzed how comorbidity networks are structured according to age groups. We conducted a retrospective study of COVID-19 mortality with hospitalized patients from Londrina, Parana, Brazil, registered in the database for severe acute respiratory infections (SIVEP-Gripe), from January 2021 to February 2022. We tested four ML models to predict the COVID-19 outcome: Logistic Regression, Support Vector Machine, Random Forest, and XGBoost. We also constructed a comorbidity network to investigate the impact of co-occurring comorbidities on COVID-19 mortality. Our study comprised 8358 hospitalized patients, of whom 2792 (33.40%) died. The XGBoost model achieved excellent performance (ROC-AUC = 0.90). Both permutation method and SHAP values highlighted the importance of age, ventilatory support status, and intensive care unit admission as key features in predicting COVID-19 outcomes. The comorbidity networks for old deceased patients are denser than those for young patients. In addition, the co-occurrence of heart disease and diabetes may be the most important combination to predict COVID-19 mortality, regardless of age and sex. This work presents a valuable combination of machine learning and comorbidity network analysis to predict COVID-19 outcomes. Reliable evidence on this topic is crucial for guiding the post-pandemic response and assisting in COVID-19 care planning and provision.

Parkinson’s disease (PD) patients are involved in motor dysfunctions and gait issues. The absence of long-term reliable gait rehabilitations could result in poor gait function, gait deficits, and locomotion problems. Meanwhile, PD patients usually face significant challenges to complete the rehabilitation programs effectively without a specialized gait laboratory or after discharge from the hospital. This paper presents a closed-loop sensing and computing system to facilitate long-term medical care for PD patients. The proposed system consists of wearable sensing, data streaming, online data processing, real-time auditory stimulation for PD gait rehabilitation, and data services for medical providers. To address the potential issues of user-friendly and comforts, we implemented a wrapped shoe based plantar pressure sensing device for under-feet pressure data collection. Afterward, quantitative measurements of gait are fulfilled through online data streaming and processing architecture. Physical therapy is generally preferred for PD rehabilitation because it provides a more significant benefit superior to the drugs. Hence, applications of gait activity recognition and closed-loop stimulation are proposed to build a reliable closed-loop system and support self-contained training. To evaluate the proposed system, we conduct experiments with three gait and PD related datasets. It broads Parkinson’s gait health care prospects that make long-term PD rehabilitation possible in community-living environment.

When monitoring the health state of people with neurological diseases, it is crucial to detect disease progression and disease-related changes as soon as possible. Such changes can be detected in the performance of everyday activities like eating, drinking, nose blowing, reading, looking at photos, knitting, telephoning, or brushing hair.

In order to supervise such everyday activities, this work proposes an approach employing smartwatches in combination with a Recurrent neural network. A smartwatch offers the possibility of integrating sensor technology into a patient’s daily routines in a unobtrusive way.

We have developed a machine learning web tool to track motion data using smartwatch sensors. Data from the accelerometer, heart rate sensor, gyroscope, gravity sensor and position sensor are tracked at 20 Hz. Our tool allows a systematic comparison of the performance between a Long Short-Term Memory (LSTM) model and the different sensor systems in classifying twelve nearly identical daily activities.

In this paper, we address the problem of caregiver effort in creating daily care documentation for dementia patients. We present an activity classification system that enables automatic classification of daily activities of patients with dementia across a nursing shift. To assist caregivers, we provide an idea for integrating our system into a nursing documentation.