ABSTRACT
Insulin pumps and other smart devices have recently made significant advancements in the treatment of diabetes, a disorder that affects people all over the world. The development of medical AI has been influenced by AI methods designed to help physicians make diagnoses, choose a course of therapy, and predict outcomes. In this article, we thoroughly analyse how AI is being used to enhance and personalize diabetes treatment. The search turned up 77 original research papers, from which we've selected the most crucial information regarding the learning models employed, the data typology, the deployment stage, and the application domains. We identified two key trends, enabled mostly by AI: patient-based therapy personalization and therapeutic algorithm optimization. In the meanwhile, we point out various shortcomings in the existing literature, like a lack of multimodal database analysis or a lack of interpretability. The rapid improvements in AI and the expansion of the amount of data already available offer the possibility to overcome these difficulties shortly and enable a wider deployment of this technology in clinical settings.
ABSTRACT
This article describes a dataset of acceleration signals acquired from a low-cost Wireless Sensor Network (WSN) during seismic events that occurred in Central Italy. The WSN consists of 5 low-cost sensor nodes, each embedding an ADXL355 tri-axial MEMS accelerometer with a fixed sampling frequency of 250 Hz. The data was acquired from February 2023 to the end of June 2023. During this period, several earthquake sequences affected the area where the sensor network was installed. Continuous data was acquired from the WSN and then trimmed around the origin time of seismic events that occurred near the installation site, close to the city of Pollenza (MC), Italy. A total of 67 events were selected, whose data is available at the Istituto Nazionale di Geofisica e Vulcanologia (INGV) Seismology data center. The traces acquired from the WSN were then manually annotated by analysts from INGV. Annotations include picking time for P and S phases, when distinguishable from the background noise, alongside an associated uncertainty level for the manual annotations. The resulting dataset consists of 328 3 × 25,001 arrays, each associated with its metadata. The metadata includes event data (hypocenter position, origin time, magnitude, magnitude type, etc.), trace-related data (mean, median, maximum, and minimum amplitudes, manual picks, and picks uncertainty), and sensor-specific data (sensor name, sensitivity, and orientation). Furthermore, a small dataset consisting of non-seismic traces is included, with the goal of providing records of noise-only traces, relative to both electronic and environmental/anthropic noise sources. The dataset holds potential for training and developing Machine Learning or signal processing algorithms for seismic data with low signal-to-noise ratios. Additionally, it is valuable for research about earthquakes, structural health monitoring, and MEMS accelerometer performance in civil and seismic engineering applications.
ABSTRACT
In response to challenging circumstances, the human body can experience marked levels of anxiety and distress. Wearable devices offer a means of real-time and ongoing data collection, facilitating personalized stress monitoring. Therefore, we collected physiological signals (blood pressure volume and electrodermal activities), using Empatica E4, from 29 subjects. A personalized protocol was developed to cause cognitive, mental, and psychological stressors since they are the ones that can be experienced in working or academic environment. We also propose a pipeline to clean and process these two signals to maximize the quality of further analysis. This study aids in the comprehension of the complex connection between stress and working situations by offering a sizable dataset made up of different physiological data. It additionally enables them to create cutting-edge stress-reduction techniques and improving professional achievement while lessening the negative impact of stress on welfare.
ABSTRACT
An effective earthquake early warning system requires rapid and reliable earthquake source detection. Despite the numerous proposed epicenter localization solutions in recent years, their utilization within the Internet of Things (IoT) framework and integration with IoT-oriented cloud platforms remain underexplored. This paper proposes a complete IoT architecture for earthquake detection, localization, and event notification. The architecture, which has been designed, deployed, and tested on a standard cloud platform, introduces an innovative approach by implementing P-wave "picking" directly on IoT devices, deviating from traditional regional earthquake early warning (EEW) approaches. Pick association, source localization, event declaration, and user notification functionalities are also deployed on the cloud. The cloud integration simplifies the integration of other services in the architecture, such as data storage and device management. Moreover, a localization algorithm based on the hyperbola method is proposed, but here, the time difference of arrival multilateration is applied that is often used in wireless sensor network applications. The results show that the proposed end-to-end architecture is able to provide a quick estimate of the earthquake epicenter location with acceptable errors for an EEW system scenario. Rigorous testing against the standard of reference in Italy for regional EEW showed an overall 3.39 s gain in the system localization speed, thus offering a tangible metric of the efficiency and potential proposed system as an EEW solution.
ABSTRACT
Smart objects and home automation tools are becoming increasingly popular, and the number of smart devices that each dedicated application has to manage is increasing accordingly. The emergence of technologies such as serverless computing and dedicated machine-to-machine communication protocols represents a valuable opportunity to facilitate management of smart objects and replicability of new solutions. The aim of this paper is to propose a framework for home automation applications that can be applied to control and monitor any appliance or object in a smart home environment. The proposed framework makes use of a dedicated messages-exchange protocol based on MQTT and cloud-deployed serverless functions. Furthermore, a vocal command interface is implemented to let users control the smart object with vocal interactions, greatly increasing the accessibility and intuitiveness of the proposed solution. A smart object, namely a smart kitchen fan extractor system, was developed, prototyped, and tested to illustrate the viability of the proposed solution. The smart object is equipped with a narrowband IoT (NB-IoT) module to send and receive commands to and from the cloud. In order to evaluate the performance of the proposed solution, the suitability of NB-IoT for the transmission of MQTT messages was evaluated. The results show how NB-IoT has an acceptable latency performance despite some minimal packet loss.
ABSTRACT
In response to challenging circumstances, the human body can experience marked levels of anxiety and distress. To prevent stress-related complications, timely identification of stress symptoms is crucial, necessitating the need for continuous stress monitoring. Wearable devices offer a means of real-time and ongoing data collection, facilitating personalized stress monitoring. Based on our protocol for data pre-processing, this study proposes to analyze signals obtained from the Empatica E4 bracelet using machine-learning algorithms (Random Forest, SVM, and Logistic Regression) to determine the efficacy of the abovementioned techniques in differentiating between stressful and non-stressful situations. Photoplethysmographic and electrodermal activity signals were collected from 29 subjects to extract 27 features which were then fed into three different machine-learning algorithms for binary classification. Using MATLAB after applying the chi-square test and Pearson's correlation coefficient on WEKA for features' importance ranking, the results demonstrated that the Random Forest model has the highest stability (accuracy of 76.5%) using all the features. Moreover, the Random Forest applying the chi-test for feature selection reached consistent results in terms of stress evaluation based on precision, recall, and F1-measure (71%, 60%, 65%, respectively).
Subject(s)
Wearable Electronic Devices , Humans , Machine Learning , Algorithms , Random Forest , Data CollectionABSTRACT
An aspect correlated with climate change is certainly represented by the alternation of severe floods and relevant drought periods. Moreover, there is evidence that changes in climate and land cover are inducing changes in stream channel cross-sections, altering local channel capacity. A direct consequence of a significant change in the local channel capacity is that the relationship between the amount of water flowing at a given point in a river or stream (usually at gauging stations) and the corresponding stage in that section, known as a stage-discharge relationship or rating curve, is changed. The key messages deriving from the present work are: (a) the more frequent and extreme the floods become, the more rapid the changes in the stream channel cross-section become, (b) from an operational point of view, the collection and processing of field measurements of the stage and corresponding discharge at a given section in order to quickly and frequently update the rating curve becomes a priority. It is, therefore, necessary to define a control system for acquiring hydrological data capable of keeping river levels and discharges under control to support flood early warnings and water management. The proposed stage-discharge management system is used by the Civil Protection Service of the Marche Region (east-central Italy) for the monitoring of river runoff in the regional watersheds. The Civil Protection Service staff performs stage-discharge field measurements using water level sensors and recorders (e.g., staff gauges, submersible pressure transducers, ultrasound and radar sensors) and a current meter, acoustic doppler velocimeter, acoustic doppler current profilers, portable mobile radar profiler and salt dilution method equipment, respectively. Power functions are fitted to the stage-discharge field data. Furthermore, extrapolation is performed to cover the full range of flow measurements; in general, extrapolation is not an easy task because of sharp changes in the stream cross-section geometry for very high or very low stages. In the present work, we also focused attention on the application problems that occur in practice and the need for frequent updating.
ABSTRACT
Natural disasters cause enormous damage and losses every year, both economic and in terms of human lives. It is essential to develop systems to predict disasters and to generate and disseminate timely warnings. Recently, technologies such as the Internet of Things solutions have been integrated into alert systems to provide an effective method to gather environmental data and produce alerts. This work reviews the literature regarding Internet of Things solutions in the field of Early Warning for different natural disasters: floods, earthquakes, tsunamis, and landslides. The aim of the paper is to describe the adopted IoT architectures, define the constraints and the requirements of an Early Warning system, and systematically determine which are the most used solutions in the four use cases examined. This review also highlights the main gaps in literature and provides suggestions to satisfy the requirements for each use case based on the articles and solutions reviewed, particularly stressing the advantages of integrating a Fog/Edge layer in the developed IoT architectures.
Subject(s)
Disasters , Internet of Things , Natural Disasters , HumansABSTRACT
The article describes a dataset of gait measures acquired to validate the use of wearable sensors in gait analysis since its measurements can be compared with those provided by the stereophotogrammetric system. The comparison with a gold standard in gait analysis makes the dataset useful for the development, testing and validation of algorithms for estimating gait parameters. The dataset contains measurements simultaneously acquired by the wearable sensors and the stereophotogrammetric system during an acquisition campaign performed on 5 healthy subjects (2 females and 3 males aged between 25 and 35 years). In the acquisition campaign the involved subjects carried out a motion task wearing the wearable sensors and reflective markers of the stereophotogrammetric system. In particular, the subjects wore in each foot a wearable sensor on the instep and a reflective marker on heel, first metatarsal head, fifth metatarsal head, and above the sensor, respectively. During the motion task each subject walked over an 11-meter long walkway according to its own course. The 5 subjects involved in the acquisition campaign performed 3 repetitions of the motion task, for a total of 15 trials in where the measures collected by wearable sensors and the stereophotogrammetric system can be compared.
ABSTRACT
The relationship between flexibility and the pattern formed by the surface electromyography activity of the back muscles while performing a dynamic trunk flexion-extension task is not yet thoroughly understood, although many previous studies have adopted it as their focus in the literature. Additionally, several studies have proposed technologies and algorithms to analyse the flexion-relaxation phenomenon, which is defined by myoelectric silence that occurs when the subject's torso exceeds a certain flexion angle. Before participating in the flexion-relaxation test, subjects involved in the data collection underwent medical examinations, in which their physical condition, perceived pain, and level of disability were reported in their anamnesis. During the flexion-relaxation test, which was conducted with 25 subjects with and without low back pain, subjects wore four surface electromyography electrodes positioned over the back muscles, as well as an inertial sensor to estimate trunk inclination.
ABSTRACT
The Industry 4.0 paradigm is based on transparency and co-operation and, hence, on monitoring and pervasive data collection. In highly standardized contexts, it is usually easy to gather data using available technologies, while, in complex environments, only very advanced and customizable technologies, such as Computer Vision, are intelligent enough to perform such monitoring tasks well. By the term "complex environment", we especially refer to those contexts where human activity which cannot be fully standardized prevails. In this work, we present a Machine Vision algorithm which is able to effectively deal with human interactions inside a framed area. By exploiting inter-frame analysis, image pre-processing, binarization, morphological operations, and blob detection, our solution is able to count the pieces assembled by an operator using a real-time video input. The solution is compared with a more advanced Machine Learning-based custom object detector, which is taken as reference. The proposed solution demonstrates a very good performance in terms of Sensitivity, Specificity, and Accuracy when tested on a real situation in an Italian manufacturing firm. The value of our solution, compared with the reference object detector, is that it requires no training and is therefore extremely flexible, requiring only minor changes to the working parameters to translate to other objects, making it appropriate for plant-wide implementation.
ABSTRACT
The article describes a dataset of doppler ultrasound audio tracks taken on a sample of 30 divers according to the acquisition protocol defined by the Divers Alert Network. The audio tracks are accompanied by a medical evaluation for the decompression sickness risk according to the Spencer's scale levels. During the acquisition campaign, each diver in the post-dive phase was subjected to a double doppler ultrasound examination of approximately 45 seconds each one in the precordial area using a Huntleigh FD1 Fetal doppler probe. The two measurements were separated by a time of 8-10 seconds necessary for carrying out specific physical exercises designed to free the bubbles trapped in the tissues. The audio tracks were stored without compression via the TASCAM DP-004 recorder and processed in order to eliminate the noise generated by the positioning of the probe and the time interval between the two measurements. The audio tracks recorded during the acquisition campaign have been evaluated by experts belonging to three independent blind teams in order to provide an assessment of the decompression sickness risk according to Extended Spencer's scale. The specific typology of doppler ultrasound audio tracks and the associated medical evaluation according to the Spencer's scale levels make this dataset useful for the development, testing, and performance evaluation of new audio processing algorithms capable of automatically detecting bubbles in the blood vessels.
ABSTRACT
The proposed dataset provides a complete set of simultaneously acquired data from contactless and wearable devices for direct and indirect heart-rate measurement. Data were acquired on a total of 20 healthy white Caucasian subjects wearing no makeup (10 males and 10 females; age: 22.50 ± 1.57 years; height: 173 ± 10 cm; weight: 62.80 ± 9.52 kg) and consisted of: i) videos of the subject's face acquired by a RGB-D (Red, Green, Blue and Depth) camera (Microsoft Kinect v2), which is a contactless device; ii) electrocardiographic (ECG) recordings acquired by a clinical Holter ECG recorder (Global Instrumentation's M12R Holter), which is a wearable device; and iii) heart-rate measurements acquired from a commercial smartwatch (Moto 360 smartwatch by Motorola), which is also a wearable device. ECG recordings were processed to extract the R-peaks position and obtain a reference indirect measurement of the heart rate. A direct measurement of the heart rate was provided by the commercial smartwatch. The dataset here presented could be useful to develop new algorithms for heart-rate detection from contactless devices and to validate contactless heart-rate estimation in comparison to reference heart rate from clinical wearable devices and to heart rate from commercial wearable devices.
ABSTRACT
This paper describes a dataset acquired on 8 subjects while simulating 13 types of falls and 5 types of Activities of Daily Living (ADL), each repeated 3 times. In details, data includes 4 simulated falls forward (falling on knees ending up lying, ending in lateral position, ending up lying, ending up lying with recovery), 4 backward (falling sitting ending up lying, ending in lateral position, ending up lying, ending up lying with recovery), 2 lateral right (ending up lying, ending up lying with recovery), 2 lateral left (ending up lying, ending up lying with recovery), and 1 syncope. Simulated ADL are: lying on a bed then standing; walking a few meters; sitting on a chair then standing; go up or down three steps; and standing after picking something. Data were acquired using a MARG sensor, a wearable multisensory device tied to the subject's waist, that recorded time-variations of the subject's acceleration and orientation (expressed through the yaw, pitch and roll angles). These data can be useful in the development and test of algorithms to automatically identify and classify fall events. Fall detection systems are particularly useful when a subject is alone and not able to stand up after a fall, since an automatic alarm can be sent remotely to receive proper help.
ABSTRACT
BACKGROUND: Tremor is one of the most common symptoms of Parkinson's disease (PD), which is widely being used in the diagnosis procedure. Accurate estimation of PD tremor based on Unified PD Rating Scale (UPDRS) provides aid for physicians in prescription and home monitoring. This article presents a robust design of a classification system to estimate PD patient's hand tremors and the results of the proposed system as compared to the UPDRS. METHODS: A smartphone accelerometer sensor is used for accurate and noninvasive data acquisition. We applied short-time Fourier transform to time series data of 52 PD patients. Features were extracted based on the severity of PD patients' hand tremor. The wrapper method was employed to determine the most discriminative subset of the extracted features. Four different classifiers were implemented for achieving best possible accuracy in the estimation of PD hand tremor based on UPDRS. Of the four tested classifiers, the Naive Bayesian approach proved to be the most accurate one. RESULTS: The classification result for the assessment of PD tremor achieved close to 100% accuracy by selecting an optimum combination of extracted features of the acceleration signal acquired. For home health-care monitoring, the proposed algorithm was also implemented on a cost-effective embedded system equipped with a microcontroller, and the implemented classification algorithm achieved 93.8% average accuracy. CONCLUSIONS: The accuracy result of both implemented systems on MATLAB and microcontroller is acceptable in comparison with the previous works.
ABSTRACT
The use of precordial Doppler monitoring to prevent decompression sickness (DS) is well known by the scientific community as an important instrument for early diagnosis of DS. However, the timely and correct diagnosis of DS without assistance from diving medical specialists is unreliable. Thus, a common protocol for the manual annotation of echo Doppler signals and a tool for their automated recording and annotation are necessary. We have implemented original software for efficient bubble appearance annotation and proposed a unified annotation protocol. The tool auto-sets the response time of human "bubble examiners," performs playback of the Doppler file by rendering it independent of the specific audio player, and enables the annotation of individual bubbles or multiple bubbles known as "showers." The tool provides a report with an optimized data structure and estimates the embolic risk level according to the Extended Spencer Scale. The tool is built in accordance with ISO/IEC 9126 on software quality and has been projected and tested with assistance from the Divers Alert Network (DAN) Europe Foundation, which employs this tool for its diving data acquisition campaigns.
ABSTRACT
The current trend in health monitoring systems is to move from the hospital to portable personal devices. This work shows how consumer devices like heart rate monitors can be used not only for applications in sports, but also for medical research and diagnostic purposes. The goal pursued by our group was to develop a simple, accurate, and inexpensive system that would use a few pieces of data acquired by the heart rate monitor and process them on a smartphone to (i) provide detailed test reports about the user's health state; (ii) store report records; (iii) generate emergency calls or SMSs; and (iv) connect to a remote telemedicine portal to relay the data to an online database. The system developed by our team uses sophisticated algorithms to detect stress states, detect and classify arrhythmia events, and calculate energy consumption. It is suitable for use by elderly subjects and by patients with heart disease (e.g., those recovering from myocardial infarction) or neurological conditions such as Parkinson's disease. Easy, immediate, and economical remote health control can therefore be achieved without the need for expensive hospital equipment, using only portable consumer devices.