Abstract

Introduction
Child morbidity and mortality in resource-limited settings is a major public health problem. The previous studies were mainly concerned with determining the prevalence of child deaths and identifying associated factors. Extracting knowledge and discovering insights from hidden patterns in child data through supervised machine learning algorithms is limited. Therefore, this study aimed to predict the under-five death of children using a best performance-supervised machine learning algorithm.

Methods
A total of 1813 samples were used from the 2019 Ethiopian Demographic and Health Survey dataset. 70% and 30% of total instances were used for training the model and measuring the performance of each algorithm with 10-fold cross-validation techniques respectively. Five supervised machine learning algorithms were considered for model building and comparison. All the included algorithms were evaluated using confusion matrix elements. Information gain value was used to select important attributes to predict child deaths. The If/then logical association was used to generate rules based on relationships among attributes using Weka version 3.8.6 software.

Results
J48 is the second-best performance algorithm next to the random forest to predict child death, with 77.8% and 93.9% accuracy, respectively. Late initiation of breastfeeding, mothers with no formal education, short birth intervals, poor wealth status of the mother, and unexposed to media were the top five important attributes to predict child deaths. A total of six associated rules were generated that could determine the magnitude of child deaths. Of these, if children were rural residents, had a short birth interval, and if born as multiples (twins), then the probability of child death was 83.6%.

Conclusions
Five machine learning algorithms were included to predict child deaths and generate rules. Hence, the random forest algorithm was the best algorithm to predict child deaths. However, the study was limited since important attributes were not included in the data source, and irrelevant values were found. So, researchers are encouraged to use machine learning algorithms for future studies including important attributes that could predict child death. The current findings would be useful for stakeholders’ preparedness, and taking proactive childcare interventions. Encouraging women in education, media access, and economic development programs are essential interventions for child death reduction.

Introduction

Under-five mortality is the most important indicator to measure the health status of children, and it is a key marker for the development of countries [1]. The under-five mortality rate is the probability of children dying before their fifth birthday [2]. Globally, nearly 44% of all under-five deaths occurred before their first month of birth [3], and an estimated 4.1 million child deaths occurred in 2017 [4]. According to the Centers for Disease Control and Prevention, child mortality in the United States in 2020 was predicted to be 5.4 deaths per 1000 live births [5].

The risk of under-five mortality is highest in low-income countries. The under-five mortality rate in low-income countries was predicted to be 69 deaths per 1000 live births in 2017, which is almost 14 times the rate in high-income countries [6,7]. In Bangladesh, 522 under-five children died per 1000 live births [8]. In 2001, under-five mortality in Nepal was projected to be 91 deaths per 1000 live births [9]. Though under-five mortality shows a reduction from 166 to 67 per 1000 live births over a period of 16 years [10], Ethiopia appears to have the fifth-highest number of new-born deaths in the world [11]. Under-five mortality is projected to cause 472,000 children to die annually in Ethiopia before their fifth birthday, which places Ethiopia sixth in the world according to the number of under-five deaths [7,12]. According to WHO 2017, more than half of under-five deaths are due to infectious diseases that are easily preventable and treatable through simple and affordable interventions [13]. Under-five mortality is also caused by undernutrition, which further leads to stunting and wasting [14].

Conclusions and recommendations

This study aimed to identify the best-supervised machine learning algorithms to classify and select important attributes to predict the death of children before their fifth birthday. In line with the objectives, six supervised machine learning algorithms were considered that accurately predict the death of children before celebrating their fifth birthday. Different confusion matrix element was used to compare the candidate-supervised machine learning algorithms. Based on the result, the random forest algorithm was the best performance model to predict the death of children before celebrating their fifth birthday. Attributes such as late initiation of breastfeeding, mothers with no formal education, the short birth interval, poor wealth status of the mother, and being unexposed to media were the top important attributes to predict child deaths.

Generating associated rules for child death was another objective of the study. Accordingly, six rules were generated that were associated with the deaths of children before celebrating their fifth birthday. The findings of this study would have practical implications by supporting policymakers and stakeholders in developing childcare intervention mechanisms and preparing themselves to care for children as early as possible. Stakeholders are recommended to encourage mothers to initiate breastfeeding at the appropriate time. Improving mothers’ wealth status, closing the gap in media access, and creating awareness among mothers would be critical interventions to enhance the survival of children in Ethiopia. The generated rules would also have theoretical implications by extracting and representing knowledge. Moreover, researchers would use this study as a baseline and framework for further research studies, including important attributes that would predict child mortality in low-income countries.

Abstract

Lymphoma and leukemia are fatal syndromes of cancer that cause other diseases and affect all types of age groups including male and female, and disastrous and fatal blood cancer causes an increased savvier death ratio. Both lymphoma and leukemia are associated with the damage and rise of immature lymphocytes, monocytes, neutrophils, and eosinophil cells. So, in the health sector, the early prediction and treatment of blood cancer is a major issue for survival rates. Nowadays, there are various manual techniques to analyze and predict blood cancer using the microscopic medical reports of white blood cell images, which is very steady for prediction and causes a major ratio of deaths. Manual prediction and analysis of eosinophils, lymphocytes, monocytes, and neutrophils are very difficult and time-consuming. In previous studies, they used numerous deep learning and machine learning techniques to predict blood cancer, but there are still some limitations in these studies. So, in this article, we propose a model of deep learning empowered with transfer learning and indulge in image processing techniques to improve the prediction results. The proposed transfer learning model empowered with image processing incorporates different levels of prediction, analysis, and learning procedures and employs different learning criteria like learning rate and epochs. The proposed model used numerous transfer learning models with varying parameters for each model and cloud techniques to choose the best prediction model, and the proposed model used an extensive set of performance techniques and procedures to predict the white blood cells which cause cancer to incorporate image processing techniques. So, after extensive procedures of AlexNet, MobileNet, and ResNet with both image processing and without image processing techniques with numerous learning criteria, the stochastic gradient descent momentum incorporated with AlexNet is outperformed with the highest prediction accuracy of 97.3% and the misclassification rate is 2.7% with image processing technique. The proposed model gives good results and can be applied for smart diagnosing of blood cancer using eosinophils, lymphocytes, monocytes, and neutrophils.

Introduction

Leukemia and lymphoma are the most frequent kinds of blood cancer in people of all ages, particularly young people. This abnormal situation is induced by red blood cell proliferation and immature growth, which can harm red blood cells, bone marrow, and the immune system [1]. Leukemia accounts for more than 3.5% of new cancer cases in the United States, with over 50,000 new cases diagnosed in 2018 [2]. Cancerous lymphoblasts in the blood travel to other organs, including the heart, brain, lungs, and arteries, before spreading to important tissues throughout the body. Red blood cells are normally in charge of transporting oxygen from the heart to all organs. They make up half of the total blood volume. White blood cells, on the other hand, serve an important role in the human immune system, serving as the first line of protection against a variety of diseases and disorders [3]. As a result, accurately identifying these white blood cells is crucial in understanding the symptoms of the problem. Their categorization is determined by their cytoplasmic composition. Changes in lymphocytes, a kind of white blood cell, cause acute lymphoblastic leukemia [4]. Acute or chronic leukemia are the two types of leukemia. Without treatment, the typical recovery period for acute myeloid leukemia is roughly three months; however, the time of appearance of chronic leukemia is longer than that of acute leukemia. Chronic lymphoblastic leukemia is the most common kind of acute leukemia, accounting for around 25% of all juvenile malignancies [5, 6]. Early detection of leukemia and lymphoma has always been difficult for researchers, clinicians, and hematologists. Leukemia symptoms include enlarged lymph nodes, paleness, fever, and weight loss, although these symptoms can also be caused by other diseases [7]. Because of the moderate nature of the symptoms, diagnosing leukemia and lymphoma in its early stages is challenging. PBS microscopic assessment is the most often used leukemia and lymphoma diagnostic approach, while the gold standard for leukemia and lymphoma diagnosis only entails obtaining and analyzing white blood cell samples [8]. Several research have utilized machine learning and deep learning and machine diagnostics approaches to laboratory image processing during the last two decades in the hopes of pushing the boundaries of late diagnosis of leukemia and lymphoma and establishing their subtypes [9]. In this research, blood smear pictures were evaluated to diagnose, distinguish, and count cells in distinct kinds of leukemia and lymphoma [10].

Conclusion

The early detection of blood cancer using white blood cells can help meritoriously in its cure. The study’s proposed framework consists of three transfer learning models AlexNet, MobileNet, and ResNet empowered with SGDM, ADAM, and RMSPROP. The proposed framework applies all transfer learning models with varying learning rates effectively to white blood cancerous cells for early classification. To enhance the results, the proposed study used image processing techniques incorporating transfer learning and achieved the highest CA of 97.3% and 2.7% MCR. All experiments in the proposed study are comprehensively explained with respect to every model training and testing phase. This study helps health 5.0 to predict blood cancer in its early stages for early treatment. Furthermore, in the future, federated machine learning using the fed average technique will play a major role in better early prediction of blood cancer empowered with fuzzed machine learning and deep learning techniques will also apply more statistical techniques such as ANOVA and Chi-Square.

Abstract

Depression is a disorder that if not treated can hamper the quality of life. EEG has shown great promise in detecting depressed individuals from depression control individuals. It overcomes the limitations of traditional questionnaire-based methods. In this study, a machine learning-based method for detecting depression among young adults using EEG data recorded by the wireless headset is proposed. For this reason, EEG data has been recorded using an Emotiv Epoc+ headset. A total of 32 young adults participated and the PHQ9 screening tool was used to identify depressed participants. Features such as skewness, kurtosis, variance, Hjorth parameters, Shannon entropy, and Log energy entropy from 1 to 5 sec data filtered at different band frequencies were applied to KNN and SVM classifiers with different kernels. At AB band (8–30 Hz) frequency, 98.43 ± 0.15% accuracy was achieved by extracting Hjorth parameters, Shannon entropy, and Log energy entropy from 5 sec samples with a 5-fold CV using a KNN classifier. And with the same features and classifier overall accuracy = 98.10 ± 0.11, NPV = 0.977, precision = 0.984, sensitivity = 0.984, specificity = 0.976, and F1 score = 0.984 was achieved after splitting the data to 70/30 ratio for training and testing with 5-fold CV. From the findings, it can be concluded that EEG data from an Emotiv headset can be used to detect depression with the proposed method.

Introduction

Depressive disorder is a highly prevalent mental illness. Sadness, loss of interest or enjoyment, feelings of guilt or low self-worth, interrupted sleep or food, fatigue, and difficulty concentrating are some characteristics of depression. It may affect a person’s capacity to operate in daily life or at work or school. According to the World Health Organization (WHO) back in 2015, almost 4.4% of the world’s population was suffering from depression [1]. Because of the COVID-19 pandemic, many people suffered from depression due to job loss, study hampering, losing close relatives, staying indoors, etc. A study showed 19.3% increase in depression symptoms among people after COVID-19 in the United States [2]. A study has shown the changes in obsession, depression, and quality of life in schizophrenia patients before and after COVID-19 [3]. When depression is severe it can lead to suicide. Every year around 800 thousand people die because of suicide [1]. In 2017, 13.2% of young adults (aged 18–25) in the U.S. suffered from depression which was 5.1% less in the year 2009 [4]. Of the deaths of young people, around 9.1% are due to suicide [5]. In most suicide cases, people had psychiatric disorders where depression is the most common disorder among others [6]. According to a recent study, insecure attachment styles are linked to greater problems such as depression, social anxiety, and suicidal thoughts [7]. So, depression is a major issue that should be diagnosed and treated at an early age to prevent suicide and for the betterment of the quality of life.

Conclusion

In this work, we have recorded EEG data of young adults (19 depressed and 13 Control) evaluated by the PHQ9 screening tool and proposed a machine learning approach to learn about the EEG properties for depression detection.

We conducted multiple experiments with the reported machine learning (SVM and KNN) classifiers with our recorded data. The first experiment we conducted was on segmentation to find the better sample length suitable for ML. From our experiments, we have identified that 5-second segments are suitable for our work. Then, we have identified a suitable frequency range from various experiments that improve performance using features that are related to depression detection. We have found out that Hjorth parameters along with Shannon entropy and long energy entropy provide better results among other reported features and the beta band (12–30 Hz) gives the highest accuracy of 97.21 ± 0.21% with 25 iterations and 5-fold CV using weighted KNN compared to the other sub-bands. By combining the sub-bands, we have also investigated some other frequency ranges. We have found out that by taking the range from alpha to beta 8–30 Hz (AB), we can improve ML performance and achieve 98.43 ± 0.15% accuracy with 25 iterations and a 5-fold CV with fine KNN classifier. Using AB (8–30 Hz), we can see a significant improvement of 1.22% accuracy and slandered deviation. To further investigate the reliability, we divided the dataset 70/30 for training and testing with 5-fold CV and 10 iterations. In this experiment, we have found out that the ML performance is better by choosing the AB (8–30 Hz) band with fine KNN classifier with an accuracy of 98.10 ± 0.11%, precision of 0.984 ± 0.003, NPV of 0.977 ± 0.002, sensitivity of 0.984 ± 0.002, specificity of 0.976 ± 0.005, and F1 score of 0.984 ± 0.001. Then, we analyzed the ML performance in different regions of the brain and concluded that using the whole brain for depression detection will give the highest accuracy. The proposed method can detect depression among young adults with minimum requirements compared to other related works.

Abstract

     The increased accuracy and exhaustivity of modern Artificial Intelligence techniques in supporting the analysis of complex data, such as medical images, have exponentially increased real-world data collection for research purposes. This fact has led to the development of international repositories and high-performance computing solutions to deal with the computational demand for training models. However, other stages in the development of medical imaging biomarkers do not require such intensive computing resources, which has led to the convenience of integrating different computing backends tailored for the processing demands of the various stages of processing workflows. We present in this article a distributed and federated repository architecture for the development and application of medical image biomarkers that combines multiple cloud storages with cloud and HPC processing backends. The architecture has been deployed to serve the PRIMAGE (H2020 826494) project, aiming to collect and manage data from paediatric cancer. The repository seamlessly integrates distributed storage backends, an elastic Kubernetes cluster on a cloud on-premises and a supercomputer. Processing jobs are handled through a single control platform, synchronising data on demand. The article shows the specification of the different types of applications and a validation through a use case that make use of most of the features of the platform.

Introduction

     Increasingly, radiology is based on objective and quantifiable data extracted from Quantitative Imaging Biomarkers (QIBs). QIBs are quantitative indicators generated from structural, functional, physiological or biological characteristics of pathological lesions [1]. In the workflow development of QIBs, complex computational functions and models automatically extract attributes, namely radiomics features, from different types of radiological images to correlate them to the phenotype or genetic signatures of the lesions. These analyses aim to early detect and classify anomalies to predict prognostics, define follow-up results, or non-invasively assess the treatment response.

     In the last years, developers have analysed the images by learning from retrospective data, enriching radiomics features with demographic, clinical, liquid biopsies and genomic data because they improve the clinical value of the biomarkers [2]. Thus, gathering data processes are crucial to developing useful Clinical Decision support Systems (CDSS) based on QIBs in clinical practice, requiring a massive storage and high-performance computing capacity [3] for managing data on image biobanks. Furthermore, the huge amount of data makes traditional statistical analyses impractical, leading to a transition to novel textitArtificial Intelligence (AI) solutions such as Deep Learning [4]. Running AI algorithms efficiently requires high-computing performance resources [5] connected to the data storage backends.

Conclusions

     This work describes the design, implementation and validation of a software architecture to support the development and application of Quantitative Image Biomarkers. It implement a federated model to synchronise data among the different storage backends linked to different processing environments, including both Cloud and HPC resources. The architecture provides a federated Authentication and Authorisation Infrastructure based on Virtual Organisations that provide coherent and scalable authorisation management across the different providers. The processing backend is supported by a Kubernetes container management platform that runs the platform services and customised applications. The architecture is the outcome of a requirement elicitation process and uses mainstream, widely available components.

      The architecture uses the abstraction of the batch, High-Throughput-Compute, High-Performance computing and Interactive jobs to provide a simplified framework to develop applications with POSIX access to the distributed storage backends. HPC Jobs are managed through a mirror Kubernetes job that interacts with HPC batch queues to provide a seamless and coherent environment.

Abstract

In conventional healthcare, real-time monitoring of patient records and information mining for timely diagnosis of chronic diseases under certain health conditions is a crucial process. Chronic diseases, if not diagnosed in time, may result in patients’ death. In modern medical and healthcare systems, Internet of Things (IoT) driven ecosystems use autonomous sensors to sense and track patients’ medical conditions and suggest appropriate actions. In this paper, a novel IoT and machine learning (ML)-based hybrid approach is proposed that considers multiple perspectives for early detection and monitoring of 6 different chronic diseases such as COVID-19, pneumonia, diabetes, heart disease, brain tumor, and Alzheimer’s. The results from multiple ML models are compared for accuracy, precision, recall, F1 score, and area under the curve (AUC) as a performance measure. The proposed approach is validated in the cloud-based environment using benchmark and real-world datasets. The statistical analyses on the datasets using ANOVA tests show that the accuracy results of different classifiers are significantly different. This will help the healthcare sector and doctors in the early diagnosis of chronic diseases.

Introduction

Healthcare has an undeniable impact on the well-being of individuals [1, 2]. Chronic diseases adversely affect the health of human beings. This requires an early and accurate diagnosis which is possible after a physical examination in the hospital. The delayed diagnosis requires patients’ admission to the hospital for treatment which results in increased healthcare cost and duration, and restricted healthcare facilities in rural and remote areas [3]. The traditional diagnosis approaches are human-centered and error-prone, e.g., missed or delayed diagnosis, failure to initiate the follow-up, inaccessible patients’ history, wrong medical prescription, and incomplete information about patient history which may lead to severe consequences to patients’ health [4]. These challenges require the healthcare domain to adopt the latest technological disruptions in the fields of IoT, artificial intelligence (AI), and ML to reduce human error.

The IoT has demonstrated its benefits in connecting medical devices, sensors, and healthcare professionals to provide quality medical services [5] which resulted in extended healthcare services and reduced costs. The medical history of patients collected through these IoT devices also helps in the effective monitoring of patients’ health and early diagnosis by healthcare service providers. These benefits have resulted in the transformation of present healthcare facilities from hospital-centric facilities to patient-centric facilities [6, 7]. The network of medical devices and healthcare professionals created through sensors is also referred to as the Internet of Medical Things (IoMT) which in a healthcare system is called healthcare Internet of Things (HIoT) [8].

Conclusion

Internet has a major impact on our lives, spanning from professional life to social relationships. The IoT has added a new dimension to this process by establishing communication among smart objects, leading to the vision of anytime, anywhere, any media, and anything communication. IoMT is a specialized version of IoT, which specifically deals with the healthcare domain and is playing a vital role in the healthcare sector to provide cost-effective solutions in terms of hospital billing and waiting time. In this paper, a new approach is presented to diagnose 6 different chronic diseases, namely, COVID-19, diabetes, pneumonia, Alzheimer, heart disease, and brain tumor, using machine learning classifiers based on 3 computing layers. The proposed approach is evaluated for each disease using different classifiers using both benchmark and real-world datasets. This also provides efficient and effective health services to the patients by disseminating the networked information to machine learning classifiers.

In the future, the proposed approach will be evaluated with more real-world datasets while considering more chronic diseases. Moreover, efforts will be made to develop cross-disciplinary chronic disease prediction models due to common feature sets.

Highlights

Abstract

Keywords

1. Introduction

2. Theoretical background

3. Conceptual model and hypothesis development

4. Methodology

5. Data analysis

6. Discussion

7. Conclusion

Authorship statement

Declaration of Competing Interest

Acknowledgements

Appendix A: Study measurement

References

Vitae

Abstract

There are numerous risks associated with the interconnection of healthcare provision and the Internet of Things (IoT), with its sensory capabilities shown to reduce confidence in novel technology due to fears of a loss of privacy. There exists a clear omission in the extant literature—consideration of gender differences in Frontline Healthcare Providers’ (FHP) behavioural intentions—which this work aims to address through the analysis of IoT-enabled healthcare applications’ (HAs) behavioural intentions in multicultural and bi-generational (Gen X, Y) context. Essentially, analysing gender and generational differences in relation to the variables (privacy, security and trust that influence risk perception; the latter alongside attitude and perceived behavioural control potentially affect the intention) affecting FHPs’ BI towards IoT enabled HAs. A novel model is presented herein, which combines Planned Behaviour (TPB), Privacy Calculus (PCT), and the trust-risk framework. Questionnaire methodology (n = 401) was applied to both generations under consideration, data was assessed using Partial Least Squares Multi-Group Analysis (PLS-MGA), which showed gender differences in Gen Y, but there was little evidence to suggest that risk perception affects any of the cohort’s behavioural intention towards the use of IoT-enabled HA, which in turn should help guide both future institutional policy and application development .

1. Introduction

Advances in network technologies, combined with mass production of smart devices equipped with sensors with continuous, bidirectional transmission, and the advent of cloud computing have been the primary drivers behind IoT development and implementation in big data driven infrastructural control (Hassan et al., 2018; Li et al., 2020; Rafique et al., 2020; Razzak et al., 2020). However, the security problems inherent in the wider internet itself remain prevalent in IoT. In truth, each element in IoT’s tri-layer structure—perception, transport, and application—requires individual consideration in that respect (Tewari and Gupta, 2020). IoT is now commonplace in modern society, often appearing in the food supply chain, logistics, mining, computing, and healthcare sectors (Pang et al., 2015; Yildirim and Ali-Eldin, 2019). Particularly in the case of the latter, the drive for service improvement has resulted in a broad body of literature considering this advance (Asif-Ur-Rahman et al., 2019; Syed et al., 2019).

The interconnected nature of the IoT enabled healthcare model (Tewari and Gupta, 2020) contains a multitude of risks relating to privacy, security and loss of trust. Limitations on the computational ability of this model means that conventional measures used to tackle security and privacy concerns often cannot be applied (Li et al., 2020), leading to the development of a concept known as the Internet of Medical Things, wherein patient data confidentiality without loss of functionality is held paramount (Li et al., 2020).

Abstract

     An individual’s health is one of the essential aspects of life, but due to the limited technology at health care centers, many people face various restrictions during their treatment. These days, Internet-of-Things (IoT) is the most fascinating topic that provides solutions to these limitations in various ways. The IoT is utilized in various healthcare conducts that include detection, treatment, and monitoring of diseases. Wearable devices are a part of IoT that is proposed for helping patients to get the correct treatment. The conventional communication networks developed for humans-based applications face many issues like stringent latency, restricted computing capability, and short battery life. On the other hand, the onset of 5G has developed a new set of technologies that offer the vital “backbone” for connecting to the billions of devices for the upcoming IoT that would completely modify our professional and private lives. Due to the data capabilities, intelligent management, and superfast connectivity of 5G, this network has enabled new health care opportunities that include treatment, data analytics, diagnostics, and imaging. In the current review, a systematic literature survey of IoT, IoT-based wearable devices, and the role of 5G in IoT for healthcare is described in detail. Furthermore, we have explained the usage of wearable devices in detecting the issue in terms of healthcare, such as curing, monitoring, and detection of disease. Nevertheless, this review article also emphasizes the employment of IoT architecture and its wearable devices in addition to the upcoming research challenges related to this area.

Introduction

     With the increase in the advancement of technology, the demand for comfort and accessibility is also increasing, leading to an increase in the development of funding programs. Usually, it has been observed that many city hospitals lack to fulfil certain medical requirements of the patients, and therefore, patients have to search in other cities/move to other countries’ hospitals to complete their desired medical needs (Surantha et al., 2021). These days, the Internet-of-Things (IoT) in healthcare has become one of the latest trending subjects and has been found as a promising characteristic to improve the current technologies that lower its cost in each aspect of healthcare. Moreover, the IoT in biomedical is considered as the need of an hour that is required to provide medical aid to humans for monitoring, managing, detecting, and taking action from the system after receiving data, thereby effectively reducing the cost of health expenses (Khan et al., 2018). IoT is the sub-category of a computer having sensors, microcontrollers, and transceivers that are interconnected to provide information to aid the user with detailed information. It is applied in various healthcare means, including detection of disease as a precaution, monitoring disease as a self-healing route, and treating disease as the healing solution (YIN et al., 2016). Wearable devices are a part of IoT that have been developed for helping people to get the correct treatment. The IoT-based wearables are smart devices that could be worn with external accessories, tattooed or even adhered on the skin, implanted in the body, or embedded in garments and clothing.

Conclusion and prospects

     With the advancement in science and technology, the accessibility of healthcare has now become easy and affordable. But, many city hospitals lack the desired medical requirements, and therefore, the introduction of biosensors has eased these problems. Moreover, these days, the Internet-of-Things (IoT) in healthcare has become one of the latest trending subjects and a promising characteristic to improve the current technologies that lower its cost in each aspect of healthcare. IoT has become the need of the hour as it is useful for monitoring, managing, detecting, and taking action from the system after receiving data, thereby effectively reducing the excess cost of health expenses. Wearable devices are a part of IoT that have been developed to help patients get the correct treatment. These IoT-based wearables devices are considered smart devices that could be worn with external accessories tattooed or even adhered to on the skin, implanted in the body, or embedded in garments and clothing. These smart devices can link with the internet for collecting, sending data, and receiving information for making smart decisions. These wearables have become a vital fragment of IoT technology, and their growth is shifting towards more specialized and practical utility from simple accessories.

Abstract

     Cardiovascular disease is a global leading cause of death, and timely monitoring can determine its extent. Clinicians use these diagnostic indicators to make scientific and reasonable decisions. However, when decision-makers (DMs) encounter risks in complex environments, their limited rationality may affect decision behaviors. Therefore, the paper explores a new three-way multi-attribute decision making method based on regret theory (3W-MADM-R), which uses heart disease data to make decisions in fuzzy environments. There are three main steps in developing 3W-MADM-R, i.e., (i) we propose the notion of relative outcome functions and corresponding aggregated regret-based utility functions of each object; (ii) we estimate the conditional probability via an outranked set defined by an outranking relation based on the Preference Ranking Organization Method for Enrichment Evaluation (PROMETHEE II); (iii) we construct three-way decision rules to solve the problems of clustering and ranking of objects in data analysis. In order to demonstrate the usefulness of 3W-MADM-R, we apply it to analyze heart disease data. By comparing with results of other methods, we show the feasibility, stability and superiority of the presented 3W-MADM-R method.

Introduction

     Multi-attribute decision making (MADM) refers to a decision problem in which the importance of different attributes is considered and the decision scheme is ranked or the optimal one is selected. There are many applications of MADM methods in management, engineering design, economics and other fields. For instance, Churchman et al. [1] first used a simple weighting method to solve the problem of “choosing the investment policy of enterprises”. At present, the classic MADM methods, including the TOPSIS method [2], the ELECTRE method [3], the PROMETHEE-II method [4] and others. Meanwhile, many types of uncertainties have been considered in MADM methods, including randomness [5], fuzziness [6] and roughness [7]. According to the behavior characteristics of a DM, stochastic, fuzzy-set, and rough-set-based MADM models have been proposed and investigated [8], [9], [10], [11], [12].

     With the rise of MADM techniques, their applications in the medical field [13] have become more and more widespread. A case in point is that the MADM method can be applied to solve cardiovascular and cerebrovascular diseases that seriously threaten human health. According to a report released by World Health Organization (WHO), 18 million people die from cardiovascular diseases each year, and 85% of which are due to heart disease and stroke. Data from the National Bureau of Statistics (https://data.stats.gov.cn/) show that the proportion of deaths due to heart disease in the total number of deaths is increasing year by year and has reached 23.65% in 2019. Therefore, the prevention and control of heart disease need joint efforts of all individuals. Early monitoring, diagnosis and treatment can effectively reduce the morbidity and mortality of heart disease and improve the life quality of patients. One of the key problems in the field of life sciences is how to quickly and effectively diagnose heart disease. At present, routine detection items for this disease mainly include blood pressure, electrocardiogram, blood routine, blood lipids, blood glucose, hemorheology, and so forth. These tests will help clinicians to determine the location and extent of cardiovascular diseases. However, in complex environments, facing with limited rationality owned by DMs and different forms of biomedical data, there are two issues that need to be addressed. One is how to effectively and reasonably deal with ambiguities in psychological and medical data along with the bounded rationality of DMs, and the other one is how DMs with limited rationality apply these data to diagnose a variety of diseases. In real world, clinical decision making and other aspects remain to be explored.

Conclusions

     In recent years, the researches on MADM problems and 3WD models have become the mainstream. On this basis, we have combined RT with the traditional PROMETHEE-II method to construct a new 3W-MADM-R method to solve the decision making problems with behavioral psychology in real life. The effectiveness and superiority of our proposed method have been verified by an illustrative example and comparative analysis of other methods. In addition, in order to demonstrate the stability and feasibility of the 3W-MADM-R method, parameter analysis has been carried out and another three data sets have been added for further verification. The main contributions of this paper are summarized as follows:

     (1) Based on the net-flow of the traditional PROMETHEE-II method, we have constructed a new outranking relation and objectively given the membership degree of the membership function to calculate the conditional probability, which solves the deficiency of the existing methods [38,39].

Abstract

1. Introduction

2. Materials and Methods

3. Proposed Epilepsy Classification Method

4. Results and Discussion

5. Conclusions

Data Availability

Conflicts of Interest

References

Abstract

     The precise detection of epileptic seizure helps to prevent the serious consequences of seizures. As the electroencephalogram (EEG) reflects the brain activity of patients effectively, it has been widely used in epileptic seizure detection in the past decades. Recently, deep learning-based detection methods which automatically learn features from the EEG signals have attracted much attention. However, with deep learning-based detection methods, different input formats of EEG signals will lead to different detection performances. In this paper, we propose a deep learning-based epileptic seizure detection method with hybrid input formats of EEG signals, i.e., original EEG, Fourier transform of EEG, short-time Fourier transform of EEG, and wavelet transform of EEG. Convolutional neural networks (CNNs) are designed for extracting latent features from these inputs. A feature fusion mechanism is applied to integrate the learned features to generate a more stable syncretic feature for seizure detection. The experimental results show that our proposed hybrid method is effective to improve the seizure detection performance in few-shot scenarios.

Introduction

     Approximately one percent of the world’s population, 65 million people, suffer from epilepsy, more than Parkinson’s disease, Alzheimer’s disease, and multiple sclerosis combined [1]. About two-thirds of people with epilepsy can be treated with medication, and the rest may require surgical intervention. Epilepsy has the characteristics of sudden and recurrent seizures, which may lead to falls, asphyxia, and even death. Therefore, seizure detection is very important for early warning and treatment of epilepsy.

     Epileptic seizure detection is mainly based on electroencephalogram (EEG) [2–4]. Single-channel EEG acquisition equipment improves the practicability of EEG in epileptic detection due to its simplicity in implementation. However, the provided information by signal-channel EEG signal is limited because of the small number of channels. Therefore, it is worth studying to establish a model with high accuracy and high robustness for single-channel EEG epileptic detection.

     The traditional methods are mainly based on feature engineering techniques which extract the corresponding features from EEG signals and then complete the detection based on the extracted features [5–9]. These features include time-domain features [10–12], frequency-domain features [8, 9], and time-frequency-domain features [13–15]. Once the features are extracted, EEG signals can be classified using a variety of classifiers. No matter what classifier is used, the quality of designed features will greatly affect the performance of epilepsy detection. In recent years, with the development of deep learning technology, many works have applied deep learning to perform epilepsy detection [16, 17]. Different from traditional feature engineering, deep learning methods automatically learn features from EEG signals and further complete detection tasks with an end-to-end manner without complicated manual feature design process and can achieve better performance than traditional methods in many scenarios.

Conclusions

     In this paper, we focus on epileptic classification in few-shot scenarios. In order to make the classification accuracy higher and more stable, we propose a deep learning method with hybrid input, i.e., original EEG signal, FFT, STFT, and DWT of EEG signal. In order to alleviate the tendency of overfitting, two means are applied. The first is that we replace the traditional convolution by depthwise separable convolution for reducing the parameters in network and then 2 regularization is applied, whose function is decreasing the complexity of the model. We conduct several experiments to distinguish normal and epileptic EEG, and the results show the proposed method with hybrid input has strong advantages in epileptic classification. It benefits from the complementarity of time-domain properties, frequency-domain properties, and time-frequency-domain properties. Our proposed method provides a new perspective to enrich the input information to make improvements for deep learning-based epileptic diagnosis.

Abstract

Study objective

     To determine whether there has been growth in publications on the use of artificial intelligence in cardiology and oncology, we assessed historical trends in publications related to artificial intelligence applications in cardiology and oncology, which are the two fields studying the leading causes of death worldwide. Upward trends in publications may indicate increasing interest in the use of artificial intelligence in these crucial fields.

Design/setting

     To evaluate evidence of increasing publications on the use of artificial intelligence in cardiology and oncology, historical trends in related publications on PubMed (the biomedical repository most frequently used by clinicians and scientists in these fields) were reviewed.

Results

     Findings indicated that research output related to artificial intelligence (and its subcategories) generally increased over time, particularly in the last five years. With some initial degree of vacillation in publication trends, a slight qualitative inflection was noted in approximately 2015, in general publications and especially for oncology and cardiology, with subsequent consistent exponential growth. Publications predominantly focused on “machine learning” (n = 20,301), which contributed to the majority of the accelerated growth in the field, compared to “artificial intelligence” (n = 4535), “natural language processing” (n = 2608), and “deep learning” (n = 4459).

Conclusion

     Trends in the general biomedical literature and particularly in cardiology and oncology indicated exponential growth over time. Further exponential growth is expected in future years, as awareness and cross-disciplinary collaboration and education increase. Publications specifically on machine learning will likely continue to lead the way.

Introduction

     Artificial intelligence (AI) is a broad concept describing computer-performed tasks that would normally require human intelligence. AI applied to medicine has impacted several disciplines, but may have the broadest applications in cutting-edge research-oriented fields such as cardiology and oncology. Cancer and cardiovascular diseases are the leading causes of death [1], [2], and may benefit from applications of AI to elucidate pathophysiology and opportunities for prevention and early diagnosis. Great evidence of growth in a field can come from acceleration in publications indicative of advancement and potential impact in the field. Yet, this has not been objectively assessed for AI publications in the general biomedical literature, nor specifically for the prominent fields of cardiology or oncology.

     AI plays a crucial role in our daily lives and our interactions with technology. There is increasing interest in the implementation of AI in healthcare. A main driver of AI in healthcare is the application of machine learning (ML) algorithms, which are a subset of AI and glean and use insights from training datasets to perform tasks and make predictions without explicit additional programming [3], [4]. Deep learning (DL) is a form of ML that is also based on pattern recognition and uses layered artificial neural networks to assess data at different levels of abstraction, to automate complex cognitive tasks that may not yet be clearly delineated [5], [6]. Natural language processing (NLP) is another subfield of AI, in which computer-based algorithms analyze, process, and transform natural language data into a form ready for computation [7], [8], [9]. All of these forms of AI have utility in healthcare, including in cardiology and oncology [10].

Results

     Publications using the term “artificial intelligence” in the title or abstract were first noted in 1963, “machine learning” in 1964, “natural language processing” in 1978, “deep learning” in general in 1989, and “deep learning” relevant to machine learning in 2011. In the last two decades, publications using all four terms related to AI showed remarkable growth (Fig. 1). Most publications used the term “machine learning” (n = 20,301) with a notable predominance (Fig. 1) compared to the other three distinct terms “artificial intelligence” (n = 4535), “natural language processing” (n = 2608), and “deep learning” (n = 4459) (Table 1). Consequently, the acceleration of publications in AI overall appeared to be due almost entirely to publications on “machine learning”.