Recognition of Daily Activities in Adults With Wearable Inertial Sensors: Deep Learning Methods Study.

BACKGROUND: Activities of daily living (ADL) are essential for independence and personal well-being, reflecting an individual's functional status. Impairment in executing these tasks can limit autonomy and negatively affect quality of life. The assessment of physical function during ADL is crucial for the prevention and rehabilitation of movement limitations. Still, its traditional evaluation based on subjective observation has limitations in precision and objectivity. OBJECTIVE: The primary objective of this study is to use innovative technology, specifically wearable inertial sensors combined with artificial intelligence techniques, to objectively and accurately evaluate human performance in ADL. It is proposed to overcome the limitations of traditional methods by implementing systems that allow dynamic and noninvasive monitoring of movements during daily activities. The approach seeks to provide an effective tool for the early detection of dysfunctions and the personalization of treatment and rehabilitation plans, thus promoting an improvement in the quality of life of individuals. METHODS: To monitor movements, wearable inertial sensors were developed, which include accelerometers and triaxial gyroscopes. The developed sensors were used to create a proprietary database with 6 movements related to the shoulder and 3 related to the back. We registered 53,165 activity records in the database (consisting of accelerometer and gyroscope measurements), which were reduced to 52,600 after processing to remove null or abnormal values. Finally, 4 deep learning (DL) models were created by combining various processing layers to explore different approaches in ADL recognition. RESULTS: The results revealed high performance of the 4 proposed models, with levels of accuracy, precision, recall, and F1-score ranging between 95% and 97% for all classes and an average loss of 0.10. These results indicate the great capacity of the models to accurately identify a variety of activities, with a good balance between precision and recall. Both the convolutional and bidirectional approaches achieved slightly superior results, although the bidirectional model reached convergence in a smaller number of epochs. CONCLUSIONS: The DL models implemented have demonstrated solid performance, indicating an effective ability to identify and classify various daily activities related to the shoulder and lumbar region. These results were achieved with minimal sensorization-being noninvasive and practically imperceptible to the user-which does not affect their daily routine and promotes acceptance and adherence to continuous monitoring, thus improving the reliability of the data collected. This research has the potential to have a significant impact on the clinical evaluation and rehabilitation of patients with movement limitations, by providing an objective and advanced tool to detect key movement patterns and joint dysfunctions.

Transparency as a Means to Analyse the Impact of Inertial Sensors on Users during the Occupational Ergonomic Assessment: A Systematic Review.

The literature has yielded promising data over the past decade regarding the use of inertial sensors for the analysis of occupational ergonomics. However, despite their significant advantages (e.g., portability, lightness, low cost, etc.), their widespread implementation in the actual workplace has not yet been realized, possibly due to their discomfort or potential alteration of the worker's behaviour. This systematic review has two main objectives: (i) to synthesize and evaluate studies that have employed inertial sensors in ergonomic analysis based on the RULA method; and (ii) to propose an evaluation system for the transparency of this technology to the user as a potential factor that could influence the behaviour and/or movements of the worker. A search was conducted on the Web of Science and Scopus databases. The studies were summarized and categorized based on the type of industry, objective, type and number of sensors used, body parts analysed, combination (or not) with other technologies, real or controlled environment, and transparency. A total of 17 studies were included in this review. The Xsens MVN system was the most widely used in this review, and the majority of studies were classified with a moderate level of transparency. It is noteworthy, however, that there is a limited and worrisome number of studies conducted in uncontrolled real environments.

Prediction of the mode of delivery using artificial intelligence algorithms.

BACKGROUND AND OBJECTIVE: Mode of delivery is one of the issues that most concerns obstetricians. The caesarean section rate has increased progressively in recent years, exceeding the limit recommended by health institutions. Obstetricians generally lack the necessary technology to help them decide whether a caesarean delivery is appropriate based on antepartum and intrapartum conditions. METHODS: In this study, we have tested the suitability of using three popular artificial intelligence algorithms, Support Vector Machines, Multilayer Perceptron and, Random Forest, to develop a clinical decision support system for the prediction of the mode of delivery according to three categories: caesarean section, euthocic vaginal delivery and, instrumental vaginal delivery. For this purpose, we used a comprehensive clinical database consisting of 25,038 records with 48 attributes of women who attended to give birth at the Service of Obstetrics and Gynaecology of the University Clinical Hospital "Virgen de la Arrixaca" in the Murcia Region (Spain) from January of 2016 to January 2019. Women involved were patients with singleton pregnancies who attended to the emergency room on active labour or undergoing a planned induction of labour for medical reasons. RESULTS: The three implemented algorithms showed a similar performance, all of them reaching an accuracy equal to or above 90% in the classification between caesarean and vaginal deliveries and somewhat lower, around 87% between instrumental and euthocic. CONCLUSIONS: The results validate the use of these algorithms to build a clinical decision system to help gynaecologists to predict the mode of delivery.

Analyzing the use of artificial intelligence for the management of chronic obstructive pulmonary disease (COPD).

OBJECTIVE: Chronic obstructive pulmonary disease (COPD) is a disease that causes airflow limitation to the lungs and has a high morbidity around the world. The objective of this study was to evaluate how artificial intelligence (AI) is being applied for the management of the disease, analyzing the objectives that are raised, the algorithms that are used and what results they offer. METHODS: We conducted a scoping review following the Arksey and O'Malley (2005) and Levac et al. (2010) guidelines. Two reviewers independently searched, analyzed and extracted data from papers of five databases: Web of Science, PubMed, Scopus, Cinahl and Cochrane. To be included, the studies had to apply some AI techniques for the management of at least one stage of the COPD clinical process. In the event of any discrepancy between both reviewers, the criterion of a third reviewer prevailed. RESULTS: 380 papers were identified through database searches. After applying the exclusion criteria, 67 papers were included in the study. The studies were of a different nature and pursued a wide range of objectives, highlighting mainly those focused on the identification, classification and prevention of the disease. Neural nets, support vector machines and decision trees were the AI algorithms most commonly used. The mean and median values of all the performance metrics evaluated were between 80% and 90%. CONCLUSIONS: The results obtained show a growing interest in the development of medical applications that manage the different phases of the COPD clinical process, especially predictive models. According to the performance shown, these models could be a useful complementary tool in the decision-making by health specialists, although more high-quality ML studies are needed to endorse the findings of this study.

Diabetic retinopathy detection through convolutional neural networks with synaptic metaplasticity.

BACKGROUND AND OBJECTIVES: Diabetic retinopathy is a type of diabetes that causes vascular changes that can lead to blindness. The ravages of this disease cannot be reversed, so early detection is essential. This work presents an automated method for early detection of this disease using fundus colored images. METHODS: A bio-inspired approach is proposed on synaptic metaplasticity in convolutional neural networks. This biological phenomenon is known to directly interfere in both learning and memory by reinforcing less common occurrences during the learning process. Synaptic metaplasticity has been included in the backpropagation stage of a convolution operation for every convolutional layer. RESULTS: The proposed method has been evaluated by using a public small diabetic retinopathy dataset from Kaggle with four award-winning convolutional neural network architectures. Results show that convolutional neural network architectures including synaptic metaplasticity improve both learning rate and accuracy. Furthermore, obtained results outperform other methods in current literature, even using smaller datasets for training. Best results have been obtained for the InceptionV3 architecture with synaptic metaplasticity with a 95.56% accuracy, 94.24% F1-score, 98.9% precision and 90% recall, using 3662 images for training. CONCLUSIONS: Convolutional neural networks with synaptic metaplasticity are suitable for early detection of diabetic retinopathy due to their fast convergence rate, training simplicity and high performance.

A DCP-based Method for Improving Laparoscopic Images.

Laparoscopy is an invasive surgical technique performed in abdominal surgery that provides faster recovery than conventional open surgeries. It requires to introduce a camera to observe the surgical maneuvers. However, during this intervention, the quality of the image may be reduced due to the creation of water vapor and carbon dioxide inside the pelvic-abdominal cavity. This phenomenon produces a nebulous image that causes interruptions during the surgical intervention. Removing this nebulous effect is a key factor to improve the vision of the surgeon. In this study, we have used a method based on the dark channel prior to remove the haze in video frames of laparoscopic surgeries to provide better quality images. The results have been positively evaluated by specialists using real video frames of laparoscopic surgeries, thus demonstrating that this method can be effective in improving the quality of the images without losing any detail of the original image.

Support System for Early Diagnosis of Chronic Obstructive Pulmonary Disease Based on the Service-Oriented Architecture Paradigm and Business Process Management Strategy: Development and Usability Survey Among Patients and Health Care Providers.

BACKGROUND: Chronic obstructive pulmonary disease (COPD) is a chronic respiratory disease with a high global prevalence. The main scientific societies dedicated to the management of this disease have published clinical practice guidelines for quality practice. However, at present, there are important weaknesses in COPD diagnosis criteria that often lead to underdiagnosis or misdiagnosis. OBJECTIVE: We sought to develop a new support system for COPD diagnosis. The system was designed to overcome the weaknesses detected in current guidelines with the goals of enabling early diagnosis, and improving the diagnostic accuracy and quality of care provided. METHODS: We first analyzed the main clinical guidelines for COPD to detect weaknesses that exist in the current diagnostic process, and then proposed a redesign based on a business process management (BPM) strategy for its optimization. The BPM system acts as a backbone throughout the process of COPD diagnosis in this proposed approach. The newly developed support system was integrated into a health information system for validation of its use in a hospital environment. The system was qualitatively evaluated by experts (n=12) and patients (n=36). RESULTS: Among the 12 experts, 10 (83%) positively evaluated our system with respect to increasing the speed for making the diagnosis, helping in interpreting results, and encouraging opportunistic diagnosis. With an overall rating of 4.29 on a 5-point scale, 27/36 (75%) of patients considered that the system was very useful in providing a warning about possible cases of COPD. The overall assessment of the system was 4.53 on a 5-point Likert scale with agreement to extend its use to all primary care centers. CONCLUSIONS: The proposed system provides a functional method to overcome the weaknesses detected in the current diagnostic process for COPD, which can help foster early diagnosis, while improving the diagnostic accuracy and quality of care provided.

Business Process Management for optimizing clinical processes: A systematic literature review.

Business Process Management is a new strategy for process management that is having a major impact today. Mainly, its use is focused on the industrial, services, and business sector. However, in recent years, it has begun to apply for optimizing clinical processes. So far, no studies that evaluate its true impact on the healthcare sector have been found. This systematic review aims to assess the results of the application of Business Process Management methodology on clinical processes, analyzing whether it can become a useful tool to improve the effectiveness and quality of processes. We conducted a systematic literature review using ScienceDirect, Web of Science, Scopus, PubMed, and Springer databases. After the electronic search process in different databases, 18 articles met the pre-established requirements. The findings support the use of Business Process Management as an effective methodology to optimize clinical processes. Business Process Management has proven to be a feasible and useful methodology to design and optimize clinical processes, as well as to automate tasks. However, a more comprehensive follow-up of this methodology, better technological support, and greater involvement of all the clinical staff are factors that play a key role for the development of its true potential.

Modelling the energy harvesting from ceramic-based microbial fuel cells by using a fuzzy logic approach.

Microbial fuel cells (MFCs) is a promising technology that is able to simultaneously produce bioenergy and treat wastewater. Their potential large-scale application is still limited by the need of optimising their power density. The aim of this study is to simulate the absolute power output by ceramic-based MFCs fed with human urine by using a fuzzy inference system in order to maximise the energy harvesting. For this purpose, membrane thickness, anode area and external resistance, were varied by running a 27-parameter combination in triplicate with a total number of 81 assays performed. Performance indices such as R2 and variance account for (VAF) were employed in order to compare the accuracy of the fuzzy inference system designed with that obtained by using nonlinear multivariable regression. R2 and VAF were calculated as 94.85% and 94.41% for the fuzzy inference system and 79.72% and 65.19% for the nonlinear multivariable regression model, respectively. As a result, these indices revealed that the prediction of the absolute power output by ceramic-based MFCs of the fuzzy-based systems is more reliable than the nonlinear multivariable regression approach. The analysis of the response surface obtained by the fuzzy inference system determines that the maximum absolute power output by the air-breathing set-up studied is 450  µ W when the anode area ranged from 160 to 200 cm2, the external loading is approximately 900 Ω and a membrane thickness of 1.6 mm, taking into account that the results also confirm that the latter parameter does not show a significant effect on the power output in the range of values studied.

A decision support system for predicting the treatment of ectopic pregnancies.

BACKGROUND AND OBJECTIVE: Ectopic pregnancy is an important cause of morbidity and mortality worldwide. An early diagnosis, as well as the choice of the most suitable treatment for the patient is crucial to avoid possible complications. According to different factors an ectopic pregnancy must be treated from surgery, using a pharmacological treatment or following a conservative treatment. In this paper, a clinical decision support systems based on artificial intelligence algorithms has been developed to help clinicians to choose the initial treatment to be followed by the patient. METHODS: A decision support system based on a three stages classifier has been developed. Each stage acts as a filter and allows re-evaluation of the classification made in the previous stage in order to find diagnostic errors. This classifier has been implemented and tested for four different aid algorithms: Multilayer Perceptron, Deep Learning, Support Vector Machine and Naives Bayes. RESULTS: The results prove that the evaluated algorithms Support Vector Machine and Multilayer Perceptron can be useful to help gynecologists in their decisions about initial treatment, especially with Support Vector Machine that presents accuracy, sensitivity and specificity outcomes about 96.1%, 96% and 98%, respectively. CONCLUSIONS: According to the results, it is feasible to develop a clinical decision support system using the algorithms that present a higher precision. This system would help gynecologists to take the most accurate decision about the initial treatment, thus avoiding future complications.

Lumbatex: A Wearable Monitoring System Based on Inertial Sensors to Measure and Control the Lumbar Spine Motion.

Measuring the curvature of the lumbar spine is an important challenge in disciplines related to physical therapy, rehabilitation, and sports medicine seeking to solve the incidence of the low back pain and other spinal disorders in the population. In clinical practice, most of the methods used are manual or depend on the trained eye of the specialist who is measuring. We have developed Lumbatex: an integrated system based on inertial sensors integrated into a wearable textile device. This device is connected via Bluetooth to software, which interprets data from the sensors and provides real-time biofeedback to users in a graphical way and also a quantitative measure of the curvature and spinal motion. The system is tested in two ways: first, checking the accuracy detecting changes in curvatures; second, evaluating the usability and comfort from the user standpoint. The accuracy is checked through a static method getting curvature values from the device placed on curved platforms and a dynamic validation with volunteers performing different exercises. The results obtained showed a high accuracy measuring changes in curvature with an error lower than 1° in the static test and good usability and comfort according to the opinion of the volunteers.

Chess Practice as a Protective Factor in Dementia.

BACKGROUND: dementia is one of the main causes of disability and dependency among the older population worldwide, producing physical, psychological, social and economic impact in those affected, caregivers, families and societies. However, little is known about dementia protective factors and their potential benefits against disease decline in the diagnosed population. Cognitive stimulating activities seem to be protective factors against dementia, though there is paucity in the scientific evidence confirming this, with most publications focusing on prevention in non-diagnosed people. A scoping review was conducted to explore whether chess practice could mitigate signs, deliver benefits, or improve cognitive capacities of individuals diagnosed with dementia through the available literature, and therefore act as a protective factor. METHODS: twenty-one articles were selected after applying inclusion and exclusion criteria. RESULTS: the overall findings stress that chess could lead to prevention in non-diagnosed populations, while little has been shown with respect to individuals already diagnosed. However, some authors suggest its capacity as a protective factor due to its benefits, and the evidence related to the cognitive functions associated with the game. CONCLUSION: although chess is indirectly assumed to be a protective factor due to its cognitive benefits, more studies are required to demonstrate, with strong evidence, whether chess could be a protective factor against dementia within the diagnosed population.

A Distributed Model for Stressors Monitoring Based on Environmental Smart Sensors.

Nowadays, in many countries, stress is becoming a problem that increasingly affects the health of people. Suffering stress continuously can lead to serious behavioral disorders such as anxiety or depression. Every person, in his daily routine, can face many factors which can contribute to increase his stress level. This paper describes a flexible and distributed model to monitor environmental variables associated with stress, which provides adaptability to any environment in an agile way. This model was designed to transform stress environmental variables in value added information (key stress indicator) and to provide it to external systems, in both proactive and reactive mode. Thus, this value-added information will assist organizations and users in a personalized way helping in the detection and prevention of acute stress cases. Our proposed model is supported by an architecture that achieves the features above mentioned, in addition to interoperability, robustness, scalability, autonomy, efficient, low cost and consumption, and information availability in real time. Finally, a prototype of the system was implemented, allowing the validation of the proposal in different environments at the University of Alicante.

Empowerment of Patients with Hypertension through BPM, IoT and Remote Sensing.

Hypertension affects one in five adults worldwide. Healthcare processes require interdisciplinary cooperation and coordination between medical teams, clinical processes, and patients. The lack of patients' empowerment and adherence to treatment makes necessary to integrate patients, data collecting devices and clinical processes. For this reason, in this paper we propose a model based on Business Process Management paradigm, together with a group of technologies, techniques and IT principles which increase the benefits of the paradigm. To achieve the proposed model, the clinical process of the hypertension is analyzed with the objective of detecting weaknesses and improving the process. Once the process is analyzed, an architecture that joins health devices and environmental sensors, together with an information system, has been developed. To test the architecture, a web system connected with health monitors and environment sensors, and with a mobile app have been implemented.

Monitoring-Based Model for Personalizing the Clinical Process of Crohn's Disease.

Crohn's disease is a chronic pathology belonging to the group of inflammatory bowel diseases. Patients suffering from Crohn's disease must be supervised by a medical specialist for the rest of their lives; furthermore, each patient has its own characteristics and is affected by the disease in a different way, so health recommendations and treatments cannot be generalized and should be individualized for a specific patient. To achieve this personalization in a cost-effective way using technology, we propose a model based on different information flows: control, personalization, and monitoring. As a result of the model and to perform a functional validation, an architecture based on services and a prototype of the system has been defined. In this prototype, a set of different devices and technologies to monitor variables from patients and their environment has been integrated. Artificial intelligence algorithms are also included to reduce the workload related to the review and analysis of the information gathered. Due to the continuous and automated monitoring of the Crohn's patient, this proposal can help in the personalization of the Crohn's disease clinical process.

Using machine learning methods for predicting inhospital mortality in patients undergoing open repair of abdominal aortic aneurysm.

An abdominal aortic aneurysm is an abnormal dilatation of the aortic vessel at abdominal level. This disease presents high rate of mortality and complications causing a decrease in the quality of life and increasing the cost of treatment. To estimate the mortality risk of patients undergoing surgery is complex due to the variables associated. The use of clinical decision support systems based on machine learning could help medical staff to improve the results of surgery and get a better understanding of the disease. In this work, the authors present a predictive system of inhospital mortality in patients who were undergoing to open repair of abdominal aortic aneurysm. Different methods as multilayer perceptron, radial basis function and Bayesian networks are used. Results are measured in terms of accuracy, sensitivity and specificity of the classifiers, achieving an accuracy higher than 95%. The developing of a system based on the algorithms tested can be useful for medical staff in order to make a better planning of care and reducing undesirable surgery results and the cost of the post-surgical treatments.

Aid decision algorithms to estimate the risk in congenital heart surgery.

BACKGROUND AND OBJECTIVE: In this paper, we have tested the suitability of using different artificial intelligence-based algorithms for decision support when classifying the risk of congenital heart surgery. In this sense, classification of those surgical risks provides enormous benefits as the a priori estimation of surgical outcomes depending on either the type of disease or the type of repair, and other elements that influence the final result. This preventive estimation may help to avoid future complications, or even death. METHODS: We have evaluated four machine learning algorithms to achieve our objective: multilayer perceptron, self-organizing map, radial basis function networks and decision trees. The architectures implemented have the aim of classifying among three types of surgical risk: low complexity, medium complexity and high complexity. RESULTS: Accuracy outcomes achieved range between 80% and 99%, being the multilayer perceptron method the one that offered a higher hit ratio. CONCLUSIONS: According to the results, it is feasible to develop a clinical decision support system using the evaluated algorithms. Such system would help cardiology specialists, paediatricians and surgeons to forecast the level of risk related to a congenital heart disease surgery.

Influence of the surrounded tissue in the detection of microcalcifications using wavelets.

Breast cancer is the most common cancer in women. Many clinical decision support systems aimed to help in the diagnosis of breast cancer have been developed because an early diagnosis is fundamental to improve the results of the treatment. Most of the developments are aimed to detect microcalcifications using the same system and parameters for all the mammograms without considering any other characteristic of the breast. In this paper we introduce the type of tissue in the breast as an element that can affect the selection of the right algorithm to improve the detection rates. We adapt the system setup depending on the type of tissue improving the results of the aid system.

Swarm intelligence applied to the risk evaluation for congenital heart surgery.

Particle Swarm Optimization is an optimization technique based on the positions of several particles created to find the best solution to a problem. In this work we analyze the accuracy of a modification of this algorithm to classify the levels of risk for a surgery, used as a treatment to correct children malformations that imply congenital heart diseases.

Growing Neural Gas approach for obtaining homogeneous maps by restricting the insertion of new nodes.

The Growing Neural Gas model is used widely in artificial neural networks. However, its application is limited in some contexts by the proliferation of nodes in dense areas of the input space. In this study, we introduce some modifications to address this problem by imposing three restrictions on the insertion of new nodes. Each restriction aims to maintain the homogeneous values of selected criteria. One criterion is related to the square error of classification and an alternative approach is proposed for avoiding additional computational costs. Three parameters are added that allow the regulation of the restriction criteria. The resulting algorithm allows models to be obtained that suit specific needs by specifying meaningful parameters.