scispace - formally typeset
Search or ask a question

Showing papers presented at "IEEE International Symposium on Medical Measurements and Applications in 2020"


Proceedings ArticleDOI
Xiaying Wang1, Michael Hersche1, Batuhan Tomekce1, Burak Kaya1, Michele Magno1, Luca Benini1 
01 Jun 2020
TL;DR: In this paper, the authors presented an accurate and robust embedded motor-imagery brain-computer interface (MI-BCI) based on EEGNet, which matches the requirements of memory footprint and computational resources of low-power microcontroller units (MCUs), such as the ARM Cortex-M family.
Abstract: This paper presents an accurate and robust embedded motor-imagery brain–computer interface (MI-BCI). The proposed novel model, based on EEGNet [1], matches the requirements of memory footprint and computational resources of low-power microcontroller units (MCUs), such as the ARM Cortex-M family. Furthermore, the paper presents a set of methods, including temporal downsampling, channel selection, and narrowing of the classification window, to further scale down the model to relax memory requirements with negligible accuracy degradation. Experimental results on the Physionet EEG Motor Movement/Imagery Dataset show that standard EEGNet achieves 82.43%, 75.07%, and 65.07% classification accuracy on 2-, 3-, and 4-class MI tasks in global validation, outperforming the state-of-the-art (SoA) convolutional neural network (CNN) by 2.05%, 5.25%, and 6.49%. Our novel method further scales down the standard EEGNet at a negligible accuracy loss of 0.31% with 7.6× memory footprint reduction and a small accuracy loss of 2.51% with 15× reduction. The scaled models are deployed on a commercial Cortex-M4F MCU taking 101 ms and consuming 4.28 mJ per inference for operating the smallest model, and on a Cortex-M7 with 44 ms and 18.1 mJ per inference for the medium-sized model, enabling a fully autonomous, wearable, and accurate low-power BCI.

39 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: It emerges that there is a significant statistical difference on six spatiotemporal parameters related to gait, turning and anticipatory postural adjustments (APA) between admission phase and discharge one, suggesting the importance of rehabilitation and of the rehabilitation institutes in the follow-up of post-surgical patients.
Abstract: Modern biomedical technologies are increasingly spreading to overcome the limitation based on the use of qualitative methods such as clinical scales, in fact in the clinical practice wearable inertial systems for gait analysis are spreading to support the clinical-therapeutic decision and to control the follow-up of patients. The purpose of this study is to investigate the rehabilitation outcome of patients under examination in terms of spatiotemporal parameters using the Stand and Walk test. A population-based sample of 30 post-surgical patients undergone hip or knee replacement surgery was studied; gender, age, weight, drug therapy and physiotherapy treatment were recorded. Data were calculated using a wearable inertial system for gait analysis: Opal System by APDM and analyzed using ANOVA test. Overall, ANOVA test between admission phase and discharge one, it emerges that there is a significant statistical difference on six spatiotemporal parameters related to gait, turning and anticipatory postural adjustments (APA). Study results suggest that there is an improvement in mobility of patients hospitalized at the Operating Rehabilitation and Functional Recovery Unit of ICS Maugeri Institute of Care and Scientific Research of Bari (Italy) after only a month of rehabilitation treatment. This result suggest the importance of rehabilitation and of the rehabilitation institutes in the follow-up of post-surgical patients.

25 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: It is proved that gait analysis and machine learning can be used to detect the presence of Mild Cognitive Impairment in PD patients through a machine learning approach.
Abstract: Parkinson Disease (PD) consists in a progressive, neurodegenerative disorder whose clinically characteristic is a combination of several motor and non-motor symptoms. Recently, the construct of Mild Cognitive Impairment (MCI), originally conceptualized to identify the pre-dementia state in Alzheimer Disease, has been employed in PD to describe a frame of cognitive decline without impaired functional activity. The aim of this study was to differentiate PD patients with and without MCI using quantitative gait variables through a machine learning approach. Thus, 45 PD patients underwent gait analysis and spatial-temporal parameters were acquired in three different conditions (normal gait, motor dual-task and cognitive dual-task). While the demographic and clinical features of PD patients with and without MCI were compared through a statistical analysis, the features of each gait condition were given as input to decision tree (DT), random forests (RF) and k nearest neighbour (KNN) to detect the presence of MCI. Then, some evaluation metrics were computed. DT achieved the highest accuracy (86.8%) using motor dual-task features, and the best sensitivity (88.2%), using gait task features as well as KNN (88.2% of sensitivity). KNN obtained the highest AUCROC (0.900) with the cognitive dual-task. DT with motor and cognitive dual-tasks and KNN with cognitive dual-task achieved the highest sensitivity (85.3%). Averaging the metrics, the cognitive dual-task showed the highest mean accuracy and specificity while the best mean sensitivity was obtained by the gait task. This paper proved that gait analysis and machine learning can be used to detect the presence in MCI in PD patients.

22 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: Study results suggested that there is a drastic change in spatiotemporal and kinematic parameters related to gait underlining how the backpack alters the latter.
Abstract: Although a very large number of students in the world use uncomfortable and heavy backpacks, their negative influence on gait in terms of spatiotemporal and kinematic parameters are still not well investigated. The purpose of the paper is to investigate the role of the school backpack during the execution of the Timed Up and Go test trying to identify if and how much it affects walking in terms of spatiotemporal and kinematic parameters considering whether it might be correlated to low back pain in children. A population-based sample of 98 children students ages 10-12 years was studied; gender, age, weight and lower limb length were recorded. Data were computed using a wearable inertial device for gait analysis: G-Walk System by BTS Bioengineering and analyzed through Inferential Statistics and Machine Learning. Overall, concerning Inferential Statistics carried out through ANOVA test for each motion parameter between free walk and walk with backpack, it was found that there is a significant statistical difference on 23 out of 30 motion parameters, of which 20 with maximum statistical significance (p<0.0001). Concerning Machine Learning analysis carried out through Random Forest algorithm considering free walk and walk with backpack as two different classes, it was found a high value of the overall Accuracy metric with a value of about 96%. Study results suggested that there is a drastic change in spatiotemporal and kinematic parameters related to gait underlining how the backpack alters the latter. The results should be taken in correct account to safeguard children’s health exposed to these prolonged condition.

21 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: Evaluated repeatability of spatio-temporal gait measurements are more repeatable than sway and anticipatory postural adjustments variables, and they are proper indexes to better identify Parkinsonian walking features.
Abstract: Parkinson’s Disease is one of the most common neurodegenerative disorders. Its principal symptoms regard motor area, and gait is one of the most affected motor characteristics. In clinical environment Parkinsonian gait is often assessed through gait analysis. Opal System by APDM is a commercial device used to perform gait analysis with inertial measurements units (IMUs). In this study we evaluate repeatability of spatio-temporal gait measurements, assessed with Opal Instrumented Stand and Walk (ISAW) test on a cohort of forty-five Parkinsonian patients. Repeatability is assessed by means of Intraclass Correlation Coefficient (ICC) and Repeatability Limit (RL) for each variable. RL is then compared to the absolute value of difference (DoM) of PD patients’ measurements mean and normative mean of the same variable, in order to understand which variable can better characterize Parkinsonian gait with respect to normal gait. Results show that gait and turn measurements are more repeatable than sway and anticipatory postural adjustments variables, and they are proper indexes to better identify Parkinsonian walking features.

20 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: Results showed that the lag introduced by the low-pass filtering of the rectified EMG, generates delays greater than those associated with the force sensor, confirming the possibility of using force sensors as a convenient alternative to EMG signals in the control of prostheses.
Abstract: Active hand prostheses are usually controlled by electromyography (EMG) signals acquired from few muscles available in the residual limb. In general, it is necessary to estimate the envelope of the EMG in real-time to implement the control of the prosthesis. Recently, sensors based on Force Sensitive Resistor (FSR) proved to be a valid alternative to monitor muscle contraction. However, FSR-based sensors measure the mechanical phenomena related to muscle contraction rather than those electrical. The aim of this study is to test the difference between the EMG and force signal in controlling a prosthetic hand. Particular emphasis has been placed on verify the prosthesis activation speed and their application to fast grabbing hand prosthesis as the "Federica" hand. Indeed, there is an intrinsic electro-mechanical delay during muscle contraction, since the electrical activation of muscle fibres always precedes their mechanical contraction. However, the EMG signal needs to be processed to control prosthesis and such filtering unavoidably causes a delay. On the contrary the force signal doesn’t need any processing. Both EMG and force signals were simultaneously recorded from the flexor carpi ulnaris muscle, while subject performed wrist flexions. The raw EMG signals were rectified and low-pass filtered to extract their envelopes. Different widespread operators were used: Moving Average, Root Mean Square, Butterworth low-pass; the cut-off frequency was set to 5 Hz. Afterward, a classic double threshold method was used to compute the muscle contraction onsets (i.e. the signal should exceed a threshold level for a certain time period). Results showed that the lag introduced by the low-pass filtering of the rectified EMG, generates delays greater than those associated with the force sensor. This analysis confirms the possibility of using force sensors as a convenient alternative to EMG signals in the control of prostheses.

17 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: The aim of this study is to validate the device, comparing its collected data to the ones obtained by standard instrument used in clinical environment, resulting that reliable measures can be obtained with the prototype device.
Abstract: SWEET Sock is a wearable prototype device designed as a sensorized sock in E-Textile, which allows the acquisition of plantar pressure and acceleration signals deriving from the motion of the lower limbs. Aim of this study is to validate the device, comparing its collected data to the ones obtained by standard instrument used in clinical environment. Accelerometric signals recorded by Sweet Sock have been compared with those obtained by the inertial system Opal by APDM, while Zebris baropodometric platform by FDM has been used as the reference system to validate pressure signals. Bland-Altman analysis has been used to assess benchmarking. Both for accelerometric and pressure signals, no systematic or proportional difference between the devices has been carried out, resulting that reliable measures can be obtained with the prototype device.

14 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: A novel automated image acquisition and analysis system combining an infrared thermal camera with a RGB-D sensor is proposed for detection of NEC in preterm newborns, demonstrating encouraging initial results.
Abstract: Necrotizing enterocolitis (NEC) is a severe condition in neonates, typically involving inflammation in the small intestine. In this paper, a novel automated image acquisition and analysis system combining an infrared thermal camera with a RGB-D sensor is proposed for detection of NEC in preterm newborns. Calibration procedures are defined to ensure frame synchronization and observation consistency among the color, depth and infrared images. Segmentation and extraction of the body are automatically performed on the infrared image with the help of the additional color and depth information collected by the RGB-D sensor. Automated analysis of the temperature distribution over the entire abdominal region minimizes interference from manual intervention and facilitates operation in a clinical environment. An experimental comparison of temperature distribution in normal babies and those with NEC demonstrates encouraging initial results.

13 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: A dynamic method based on Compressed Sensing to reconstruct multi-lead electrocardiography signals in support of Internet-of-Medical-Things by dynamically evaluated through the signal samples acquired by the first lead.
Abstract: This paper proposes a dynamic method based on Compressed Sensing (CS) to reconstruct multi-lead electrocardiography (ECG) signals in support of Internet-of-Medical-Things. Specifically, the sensing matrix is dynamically evaluated through the signal samples acquired by the first lead. The experimental evaluation demonstrates that, compared to the traditional CS multi-lead method adopting a random sensing matrix, the proposed dynamic method exhibits a lower difference from the original ECG signal.

12 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: Results reveal that the model built by the approach proposed in this paper can effectively achieve a better prediction and enhance the accuracy of the model, which proves that K-Fold can achieve better performance on a small skin cancer dataset.
Abstract: Deep convolution neural networks (DCNNs) enable effective methods to predict the melanoma classes otherwise found with ultrasonic extraction. However, gathering large datasets in local hospitals in Sweden can take years. Small datasets will result in models with poor accuracy and insufficient generalization ability, which has a great impact on the result. This paper proposes to use a K-Fold cross validation approach based on a DCNN algorithm working on a small sample dataset. The performance of the model is verified via a Vgg16 extracting the features. The experimental results reveal that the model built by the approach proposed in this paper can effectively achieve a better prediction and enhance the accuracy of the model, which proves that K-Fold can achieve better performance on a small skin cancer dataset.

12 citations


Proceedings ArticleDOI
01 Jun 2020
TL;DR: This study proved that machine learning can help to identify patients affected by mild form of FoG that exposes them to a major risk of developing more severe form of freezing with a consequent increased risk of falling.
Abstract: Parkinson’s Disease (PD) is a common neurodegenerative disorder whose clinical picture is characterized by motor and non-motor symptoms. One of motor symptoms is freezing of gait (FoG) that consists in a few seconds during which patients can't start to walk again. In this paper 41 patients affected by PD, with and without FoG, underwent gait analysis performing three gait tasks: normal gait, a motor dual task and a cognitive task. A statistical analysis was performed on clinical, demographical and on the spatial and temporal parameters in order to find any difference between PD patients with and without FoG; the last one obtained no statistically significant results. Thus, a machine learning analysis was implemented employing tree-based algorithms (decision tree, Random Forests, Gradient Boosted Tree, Ada-Boosting of a decision tree) and using as input the spatial and temporal features of gait. The results were promising since accuracy, specificity and sensitivity overcame 90%, reaching also 100% of sensitivity in some cases. The best algorithms were Gradient Boosted Tree and the Ada-Boosting of a decision tree while Random Forests and decision tree obtained lower results. This study proved that machine learning can help to identify patients affected by mild form of FoG that exposes them to a major risk of developing more severe form of freezing with a consequent increased risk of falling.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: A hardware design model of a machine learning based fully connected neural network for detection of respiratory failure among neonates in the Neonatal Intensive Care Unit (NICU) and the proposed design methodology enables the simplification of the models for future low-cost neural network- on-chip hardware implementation.
Abstract: This paper proposes a hardware design model of a machine learning based fully connected neural network for detection of respiratory failure among neonates in the Neonatal Intensive Care Unit (NICU). The model has been developed for a diagnostic system comprised of pyroelectric transducer based breathing monitor and a pulse oximeter that can detect apneic events. The input signal of the proposed system is a digitally converted sensory data from the sensors which is processed using machine learning model to detect if apnea condition has occurred in the patient. The accuracy rate of the proposed model is around 99 percent. The proposed design methodology enables the simplification of the models for future low-cost neural network- on-chip hardware implementation.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: Evaluating the cardiac activity in children - with and without neurodevelopmental disorders - through a novel wearable solution in order to compare the stress response in different structured activities and games highlighted how wearable devices could help in estimating stress indicators for long-tests.
Abstract: Monitoring physiological parameters under stress conditions – i.e., heart rate, breath frequency or heart rate variability - through non-invasive and comfortable wearable devices was a research topic of great interest in many medical fields. In the last decade, several wearable devices and methods have been developed for stress monitoring, showing suitable performance in estimating the stress indicator. In spite of the interest in the field, the development of wearable solutions suitable for child neuropsychiatry applications was still an open challenge. In this study, we evaluate the cardiac activity in children - with and without neurodevelopmental disorders - through a novel wearable solution in order to compare the stress response in different structured activities and games. Each subject was equipped with a 3-lead electrocardiograph device and a piezoelectric respiratory sensor embedded into a thoracic belt. Subjects were asked to carry out five different activities previously chosen from a subset of specific behavioral tests (three different free-games and two structured activities). All experimental sessions were video-recorded. Results highlighted how wearable devices could help in estimating stress indicators for long-tests (over twenty-five minutes). The clinical application was conducted on a cohort of 32 children so divided: 13 with Specific Language Disorder, 15 with Autism Spectrum Disorder, and 4 control children without neurodevelopmental disorders. Statistical differences were observed between populations in the heart rate. Moreover, all subjects experienced stress effects evidenced by variation of heart rate and standard deviation, which are supported by the video analysis.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: Some of the issues associated with utilizing audio spectrograms to retrain the AlexNet image classifier for the purpose of remote patient monitoring are described and the spatial invariance assumption of the classifier is investigated.
Abstract: Image classification has had huge success in recent years, mainly due to the vast array of databases available. The lack of audio databases presents a problem when it comes to creating a deep neural network classifier aimed at measurement and monitoring of health-related sounds. Such sounds (i.e. cough) can be indicative of worsening health conditions, specifically as it relates to remote monitoring of older adults. The application of pre-existing deep neural network image classifiers to audio classification has been presented as a potential solution. This paper describes some of the issues associated with utilizing audio spectrograms to retrain the AlexNet image classifier for the purpose of remote patient monitoring. The spatial invariance assumption of the classifier is further investigated by creating two different classification tasks based on spectrograms computed from notes on a classical piano at four different noise levels; (1) octave classification and (2) note classification. As expected, the AlexNet classifier with clean data performs better when classifying octaves (98%), when compared to the note classification (83 %). When evaluating on audio with noise, the note classifier performance decreases more than the octave classification performance.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: A two-dimensional markerless clinical gait analysis protocol to estimate the sagittal lower limb joint kinematics from the markerless recordings of a single RGB-Depth camera is presented and proposed to use to develop a new generation of low-cost movement analysis systems.
Abstract: This work presents a two-dimensional markerless clinical gait analysis protocol to estimate the sagittal lower limb joint kinematics from the markerless recordings of a single RGB-Depth camera. The proposed method includes a subject separation from the background, the definition of a multi-segmental model of the lower limb and the estimation of the relevant joint kinematics. The segmentation algorithm performance was assessed by measuring the similarity between the computer-obtained segmentations and manual tracings (ground-truth). The estimated joint angles were compared to those obtained using a reference optoelectronic marker-based clinical protocol. The offset between the mean waveforms and the RMS value of the waveforms difference after removing their offset were computed. The segmentation accuracy resulted to be higher than 0.92 and very repeatable (STD of JI about 0.01). The RMSD values of the ankle kinematics (3.4° on average) are lower than those of other joints (4.9° for the hip joint and 6.2° for the knee joint, on average). Overall, given the good agreement between our results and those of marker-based method, we propose to use it to develop a new generation of low-cost movement analysis systems.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: This paper presents the experimental performance assessment of a smartwatch measuring the heart rate variability (HRV), compared to a multi-parametric chest belt that is considered as a reference sensor, indicating that at rest condition the HRV measured directly with the onboard system of the SW can be used to assess correctly theHRV.
Abstract: This paper presents the experimental performance assessment of a smartwatch (SW) measuring the heart rate variability (HRV), compared to a multi-parametric chest belt that is considered as a reference sensor. HRV from smartwatch can be extracted with two methods: directly from internal onboard processing of the device or by post-processing data collected from the photoplethysmography signal. To evaluate the uncertainty of both methods, measurements were performed while users were sitting at rest, wearing the SW on the preferred wrist and a chest belt that can collect electrocardiographic signal, used as reference measurement. Measurements from SW and belt were compared turning out to provide that HRV measured with the SW (onboard processing) has an uncertainty of 0.95% with a coverage factor k = 2, corresponding to ± 4 ms; while HRV extracted from the PPG signal has an uncertainty of 1.2%, corresponding to ± 6 ms, indicating that at rest condition the HRV measured directly with the onboard system of the SW can be used to assess correctly the HRV.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: Preliminary experimental results clearly show the capability of the realized system to accurately estimate both the tremor frequency and the hand trajectory, even if high levels of tremor are applied.
Abstract: The paper proposes a first step toward the realization of a low cost and scalable clinical device to monitor the evolution of Parkinson’s disease based on magnetic measurements. The main aim of the overall research project is the realization of an Internet-connected device able to allow the execution of standardized tests and to share results with specialized medical centers. At this stage, the system has been focused on two parameters useful for medical exams: tremor and hands trajectory. This paper describes the realized set-up based on fixed receiving coils and one mobile transmitting coil. It reports preliminary experimental results based on a robotic arm adopted as a reference. Obtained results clearly show the capability of the realized system to accurately estimate both the tremor frequency and the hand trajectory, even if high levels of tremor are applied.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: A novel method for monitoring the flow rate in IV infusions is presented, that is based on deep learning computer vision techniques, that can be effectively adopted to implement monitoring and control systems for health facilities.
Abstract: Intravenous (IV) infusion is one of the most common therapies in hospitalized patients. Monitoring the flow rate of the fluid that is being administered to the patient is therefore very important for his safety, considering that both over-infusion and under-infusion can cause serious health problems. In this document, a novel method for monitoring the flow rate in IV infusions is presented, that is based on deep learning computer vision techniques. Basically, the drip chamber is filmed with a camera and object detection is used to count drops. The proposed method is therefore less invasive than other ones developed for this purpose. Experimental results show that it can produce an accurate real-time estimate of the instantaneous flow rate of the drip. For these reasons, the proposed method can be effectively adopted to implement monitoring and control systems for health facilities.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: Variations in stress levels of the employees before, during and after lunch were studied to understand how food affects stress and the decrease in Stress levels observed shows the impact of food intake in coping with stress.
Abstract: Workplace Stress can impact the employees especially when their perception of job demands are more than what they can meet. Excessive and long term stress can have detrimental effects on both physiological and mental health. Building resilience is necessary for an individual, which helps combat the mental strain of work stress and exercise control over the work demands. At their core, stress responses are often linked to resolving the threats by increasing an individual's ability to cope. Unobtrusive and continuous monitoring of stress during work can help in understanding how stress levels change throughout a work day. In this study, stress has been evaluated from Electrocardiogram (ECG) derived Heart Rate Variability (HRV) obtained using an unobtrusive chest wearable. A total of 85 employees participated in the study and their measured stress levels were classified into four different levels as no, low, medium and high stress. The stress levels were evaluated every 5 minutes as the employees engaged in tasks like review meeting and accomplishing deadlines. Analysis revealed stress levels of employees reduced when they took breaks during work time. Females generally experienced high stress levels for much longer duration than males, whereas males experienced no, low and medium stress only a little longer than females. Also, variations in stress levels of the employees before, during and after lunch were studied to understand how food affects stress and the decrease in stress levels observed shows the impact of food intake in coping with stress.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: Although preliminary, EEG power results suggest that mental workload associated with AR usage may derive from enhanced difficulty associated with the task, and this work quantifies the reduction of users’ performance based on starting and end points gap errors.
Abstract: Augmented reality (AR) head mounted displays (HMDs) combine user’s natural view of the real world with virtual data. Thus, they might be particularly suited to guide manual tasks as in computer-aided-surgery. However, the typical focal plane of commercial devices is outside the user’s peripersonal space (i.e. the space containing reachable objects), limiting the performance of manual-task guidance. Specifically, known issues such as the "vergence-accomodation-conflict" and the "focus-rivalry" may lead to visual fatigue and mental workload worsening task performance. Here, we exploit EEG recordings during a "connecting-the-dots" task performed with and without AR to study the effects of mental workload associated with AR-related visual fatigue. First, we quantify the reduction of users’ performance based on starting and end points gap errors. Then, we investigate the effects on AR usage on cortical activity through the analysis of EEG power and Frontal Alpha Asymmetry (FAA) index. Although preliminary, EEG power results suggest that mental workload associated with AR usage may derive from enhanced difficulty associated with the task.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: A new CP is designed and fabricated along with a graphical user-friendly interface program integrally called "QCT" enabling to measure IAEA/ACR-based standard image parameters and beyond metrics including CT calibration curve, CT number of multiple objects, contrast-to-noise ratio, the edge spread function, the line spread function and the modulation transfer function.
Abstract: Computed Tomography (CT) is one of the most widely used screening and diagnostic tools in medical imaging centers. Considering IAEA HUMAN HEALTH SERIES No. 19 and the American College of Radiology (ACR) Accreditation Program, quality assurance (QA) and quality control (QC) are mandatory programs to periodically monitor the system condition to promote the effective utilization of ionization radiation for a diagnostic outcome through obtaining and retaining appropriate image quality and reduction of patient dose. Computational phantoms (CPs) are the key tool to monitor system condition. The commercial QC phantoms are expensive products and are not flexible enough for user demands. In this paper, we designed and fabricated a new CP along with a graphical user-friendly interface program integrally called "QCT" enabling to measure IAEA/ACR-based standard image parameters and beyond metrics including CT calibration curve, CT number of multiple objects, contrast-to-noise ratio, the edge spread function, the line spread function, the modulation transfer function, spatial resolution, noise power spectrum, image noise, and uniformity. The experimental assessment of QCT was tested on a GE LightSpeed VCT multi-detector CT scanner available in our clinic. In addition, we reported the details of fabrication process of our QC phantom, enabling readers to create flexible and affordable QC phantoms.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: In this work, the well-known and widespread LoRaWAN communication infrastructure is evaluated for possible complementing indoor and outdoor positioning solutions, in order to alert a remotely-connected caregiver about accidents (e.g., falls).
Abstract: It is common for elderly to experience a decrease in health conditions, limiting independence. Very often, hospitalization is not required and at home assistance is a more effective solution. Ambient assisted living technologies can help in mitigating the need for continuous supervision, enabling the elderly to easily look for help in case of emergency. In this work, the well-known and widespread LoRaWAN communication infrastructure is evaluated for possible complementing indoor and outdoor positioning solutions, in order to alert a remotely-connected caregiver about accidents (e.g., falls). The choice of LoRaWAN is dictated by the capability to implement both private and public infrastructures. Satellite-based systems are addressed for outdoor localization, whereas the demanding task of indoor localization is solved by means of Ultra Wide Band technology. Results demonstrated sub-meter error in a typical indoor scenario, with average communication latency on the order of 700 ms.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: A vision-based system made of three low-cost cameras, able to automatically infer important mobility parameters by observing the execution of well-established tests for stability assessment, and trained to predict the risk of fall of patients within 5 classes of interest.
Abstract: Falls represent one of the most serious clinical problems in the elderly population. This risk is even more important in people suffering from neurodegenerative problems. This work aims to instrumentally assess the balance performance of elderly people and specifically those suffering from neurodegenerative diseases, to obtain an objective evaluation of their risk of falls. This paper presents a vision-based system made of three low-cost cameras, able to automatically infer important mobility parameters by observing the execution of well-established tests for stability assessment. This result is achieved by a dedicated image processing pipeline, which processes videos to get dynamic user skeletons, and the following strategy for information management, which targets to feature extraction. This information finally feeds a classifier, namely a decision tree, trained to predict the risk of fall of patients within 5 classes of interest. Actual experiments performed on actual video recordings prove a good agreement of results with those expected, labeled by expert therapists, with final prediction accuracy of 79.1%.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: Compared to a static ROI, the proposed semi-automated method achieves significantly improved tracking of the patient’s face, as demonstrated by an area under the curve > 0.63 across all patients.
Abstract: Noncontact video-based patient monitoring promises several advantages over wearable sensors, particularly for patients in the NICU who have fragile skin. However, such approaches often require definition of a region-of-interest (ROI), such as the patient’s forehead. For example, a number of neonatal monitoring studies have estimated heart rate and respiration from video by first manually cropping the face of the patient before performing analyses within that region. Relying on a static ROI can fail due to patient motion or during clinical interventions, thereby demanding additional manual ROI selection over the course of the monitoring period. Widely used face detection algorithms tend to fail in a neonatal context. We therefore propose a semi-automated method where the ROI is automatically and repeatedly reinitialized to ensure robustness of the ROI for continuous monitoring. Factors such as the displacement of the patient and the change in patient poses are addressed using multiple computer vision techniques before selecting a comprehensive method for ROI tracking. Results were obtained from three patients admitted at the NICU using 20-minute videos including periods of rest, motion, and occlusion events. Compared to a static ROI, the proposed method achieves significantly improved tracking of the patient’s face, as demonstrated by an area under the curve > 0.63 across all patients.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: A DAQ based system supervised by a LabVIEW Virtual Instrument to estimate the HR by means of an accelerometric plethysmograph to monitor the patient's heartbeat at rest by identifying possible pathological conditions or diseases is developed.
Abstract: The increasing need to monitor Heart Rate (HR) has led to the development of dedicated devices based on miniaturized sensors and to their consequent pervasiveness. In this paper the aim is to develop a DAQ based system supervised by a LabVIEW Virtual Instrument to estimate the HR by means of an accelerometric plethysmograph. Two different signal processing techniques have been developed to estimate the HR respectively in time- and frequency domain, to study and compare their performances and to monitor the patient's heartbeat at rest by identifying possible pathological conditions or diseases.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: The proposed work describes preliminary results of a research project based on the realization of a Decision Support System -DSS- platform embedding medical and artificial intelligence -AI- algorithms suitable for the optimization of assistance processes of patients affected by Monoclonal Gammopaty.
Abstract: The proposed work describes preliminary results of a research project based on the realization of a Decision Support System -DSS- platform embedding medical and artificial intelligence -AI- algorithms. Specifically the telemedicine platform is suitable for the optimization of assistance processes of patients affected by Monoclonal Gammopaty. The results are related to the whole design of the platform implementing a DSS based on a multi-level decision making process. Starting from the main architecture specifications, is formulated a flowchart based on different alerting levels of patient risk including artificial intelligence -AI- decision supporting facilities. Finally, the perspectives of the performed research are discussed.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: The possibility to realize a wearable, compact remote sensor, totally passive, fully integrated with an antenna, based on a Surface Acoustic Wave resonator designed to operate at 800 MHz and realized on a flexible and biocompatible polymeric substrate, made of Polyethylene naphthalate.
Abstract: Acoustic wave devices are an attractive technology for use in sensors since acoustic waves present high sensibilities to external parameters in terms of phase velocity and damping. This technology is very interesting in environments such as Internet of Things, where low power consumption is a central requirement. The main drawback of this technology concerns the application and the reliability of the RF signal powering the sensor that makes necessary the use of virtual network analyser or spectrum analyser. Although in literature examples of acoustic wave devices used with external antennas have been discussed, to the best of our knowledge, at the state of the art acoustic devices integrated with an antenna have not been reported yet. This paper will discuss the possibility to realize a wearable, compact remote sensor, totally passive, fully integrated with the antenna. The wearable sensor is based on a Surface Acoustic Wave (SAW) resonator designed to operate at 800 MHz and realized on a flexible and biocompatible polymeric substrate, made of Polyethylene naphthalate (PEN). The SAW resonator consists of a pair of reflecting gratings, defining the acoustic cavity, and an interdigital transducer (IDT) placed at the centre of the cavity. The distributed feedback cavity shows a high Qfactor Q ≈ 2×105 when 200 reflectors are considered.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: An instrument based on the integration of Brain Computer Interface (BCI) and Augmented Reality (AR) is proposed for robotic autism rehabilitation and provides positive feedback on device acceptance and attentional performance.
Abstract: An instrument based on the integration of Brain Computer Interface (BCI) and Augmented Reality (AR) is proposed for robotic autism rehabilitation. Flickering stimuli at fixed frequencies appear on the display of Augmented Reality (AR) glasses. When the user focuses on one of the stimuli a Steady State Visual Evoked Potentials (SSVEP) occurs on his occipital region. A single-channel electroencephalographic Brain Computer Interface detects the elicited SSVEP and sends the corresponding commands to a mobile robot. The device’s high wearability (single channel and dry electrodes), and the trainingless usability are fundamental for the acceptance by Autism Spectrum Disorder (ASD) children. Effectively controlling the movements of a robot through a new channel enhances rehabilitation engagement and effectiveness. A case study at an accredited rehabilitation center on 10 healthy adult subjects highlighted an average accuracy higher than 83%. Preliminary further tests at the Department of Translational Medical Sciences of University of Naples Federico II on 3 ASD patients between 8 and 10 years old provided positive feedback on device acceptance and attentional performance.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: This work demonstrates how meditation, concentration, and training exercises can be considered as excellent forms of training for the reduction of impulsiveness and risk appetite, showing their feasibility in the contexts of intervention and cure for addictions.
Abstract: Impulsivity represents one of the risk factors strongly related to dependent behavior in subjects with a diagnosis of substance dependence. The risk appetite, in particular, represents a construct connected with other components such as jumping to conclusion and risk taking.There are several forms of treatment used with subjects who have problems related to impulsivity. A good part of these is aimed at subjects with ADHD (Attention Deficit Hyperactivity Disorder), in the form of training or exercises implemented on software. This work aims to evaluate the use of training for concentration, attention and meditation on reducing risk appetite. The experimental design, of a pre-post type, involves assessing the risk appetite before and after training, to be carried out in six sessions over two weeks. The comparison between the experimental group and the control groups showed significant efficacy in reducing the levels of impulsivity in subjects with addiction.This work represents a pilot study on the possible use of meditation, concentration, and training exercises and demonstrates how these can be considered as excellent forms of training for the reduction of impulsiveness and risk appetite, showing their feasibility in the contexts of intervention and cure for addictions.

Proceedings ArticleDOI
01 Jun 2020
TL;DR: The proposed wearable device embedding a conductive textile element seems to be suitable for wrist motions monitoring with promising applications in contexts as rehabilitation, sports, and clinical research.
Abstract: Wearable strain sensors potentially hold great significance in several areas, such as rehabilitation, research, sports. Conductive textile elements are excellent candidates for the development of wearable devices as human motion capture systems because of their intrinsic properties of lightweight, compliance, flexibility, and stretchability. Herein, the metrological characterization of sensing elements based on a conductive textile intended for joint motion monitoring is presented.The aim of this study is twofold: (i) analyzing the influence of the shape on the metrological properties of conductive textiles, (ii) providing a preliminary assessment of the performance of a custom-made wearable device (i.e., a glove) embedding a conductive textile element to monitor wrist joint movements. The static characterization and the hysteresis analysis were performed on two samples of conductive textile elements, i.e., the zig-zag pattern and the rectangular one. The feasibility assessment of a glove embedding one sensing element for wrist flexion-extension monitoring was performed on a single volunteer.Results showed that the zig-zag patterned conductive element had a resistance change range comparable to the rectangular one (i.e., 100 kΩ – 170 kΩ vs. 40 kΩ – 80 kΩ), higher averaged sensitivity (i.e., -3.96 kΩ•%-1 vs. -2.42 kΩ•%-1), and a lower hysteresis error, 6.83%). the % (i.e., averaged 5.22% vs. Regarding feasibility assessment of %the wearable device, results showed good performance in tracking wrist flexion-extension in real-time. The proposed wearable device embedding a conductive textile element seems to be suitable for wrist motions monitoring with promising applications in contexts as rehabilitation, sports, and clinical research.