A Deep Learning based IoT Framework for Assistive Healthcare using Gesture Based Interface

Around the world, the number of senior citizens is increasing and shall continue to increase, and it is expected to be around 20 percent by 2050. Realizing its importance, the United Nations has identified Health and Wellness as one of the Sustainable Development Goals (SDG). The unfortunate pandemic situation due to the COVID-19 outbreak opened up new challenges for contact-less interactions and control of devices for ensuring the well being of citizens. In this paper, our main aim is to develop an intelligent framework based on a gesture-based interface that will help the senior citizens and physically challenged people interact and control different devices using only gestures. We focus on dynamic gesture recognition using a deep learning-based Convolutional Neural Network (CNN) model. The proposed system records continuous real-time data streams from non-invasive wearable sensors. This real-time continuous data stream is fragmented into data segments that are most likely to contain meaningful gesture data frames using the Adaptive Threshold Setting algorithm. The segmented data frames are provided as input to the CNN model to train, test, validate, and then classify it into predefined clusters, which are gestures. We have used the MPU6050 Inertial Measurement Unit based sensor model for collecting the data of the hand/ finger movement. The popular and widely used ESP8266 controller is used for data gathering, processing, and communicating. We created a dataset for 36 gestures, which includes ten digits and 26 English alphabets. For each gesture, a dataset of 300 samples has been created from 5 subjects of age group between 21-30. Thus, the final dataset consists of a total of 10800 samples belonging to 36 gestures. A total of six features comprising linear accelerations and angular rotation in 3-dimensional axes are used for training and validation. The proposed model can segment 93.75% of data segments correctly using the adaptive threshold selection algorithm, and the CNN classification algorithm can classify 98.67% gestures correctly. © 2022 IEEE.

Implementing Inertial Measurement Unit on Covid Robot with Complementary and Madgwick Filter

Rapidly developing technology causes autonomous robots to develop significantly as well. This causes the demand for autonomous robots to increase in various industrial sectors ranging from agriculture, large-scale manufacturing industries, and even hospitals. Especially in the midst of the Covid-19 pandemic where physical distancing is applied to make autonomous robots that can be used as a substitute for medical personnel. The Inertial Measurement Unit (IMU) is a very important part of an autonomous robot because the IMU can measure 3 axes. The IMU sensor has been integrated with 3 other sensors, namely accelerometer, gyroscope, and magnetometer sensors. However, the data obtained from the sensor has an error value that can cause noise. Therefore, a filter is needed to get high accuracy results. In this study, complementary methods and Madgwick filters were used to reduce noise in the raw data so that the results can be maximized.

Human Movement Quality Assessment Using Sensor Technologies in Recreational and Professional Sports: A Scoping Review.

The use of sensor technology in sports facilitates the data-driven evaluation of human movement not only in terms of quantity but also in terms of quality. This scoping review presents an overview of sensor technologies and human movement quality assessments in ecologically-similar environments. We searched four online databases to identify 16 eligible articles with either recreational and/or professional athletes. A total of 50% of the studies used inertial sensor technology, 31% vision-based sensor technology. Most of the studies (69%) assessed human movement quality using either the comparison to an expert's performance, to an exercise definition or to the athletes' individual baseline performance. A total of 31% of the studies used expert-based labeling of the movements to label data. None of the included studies used a control group-based study design to investigate impact on training progress, injury prevention or behavior change. Although studies have used sensor technology for movement quality assessment, the transfer from the lab to the field in recreational and professional sports is still emerging. Hence, research would benefit from impact studies of technology-assisted training interventions including control groups as well as investigating features of human movement quality in addition to kinematic parameters.

Feedback on Trunk Movements From an Electronic Game to Improve Postural Balance in People With Nonspecific Low Back Pain: Pilot Randomized Controlled Trial.

BACKGROUND: Postural balance is compromised in people with low back pain, possibly by changes in motor control of the trunk. Augmenting exercising interventions with sensor-based feedback on trunk posture and movements might improve postural balance in people with low back pain. OBJECTIVE: We hypothesized that exercising with feedback on trunk movements reduces sway in anterior-posterior direction during quiet standing in people with low back pain. Secondary outcomes were lumbar spine and hip movement assessed during box lift and waiter bow tasks, as well as participant-reported outcomes. Adherence to the exercising intervention was also examined. METHODS: A randomized controlled trial was conducted with the intervention group receiving unsupervised home exercises with visual feedback using the Valedo Home, an exergame based on 2 inertial measurement units. The control group received no intervention. Outcomes were recorded by blinded staff during 4 visits (T1-T4) at University Hospital Zurich. The intervention group performed 9 sessions of 20 minutes in the 3 weeks between T2 and T3 and were instructed to exercise at their own convenience between T3 and T4. Postural balance was assessed on a force platform. Lumbar spine and hip angles were obtained from 3 inertial measurement units. The assessments included pain intensity, disability, quality of life, and fear of movement questionnaires. RESULTS: A total of 32 participants with nonspecific low back pain completed the first assessment T1, and 27 (84%) participants were randomized at T2 (n=14, 52% control and n=13, 48% intervention). Intention-to-treat analysis revealed no significant difference in change in anterior-posterior sway direction during the intervention period with a specified schedule (T2-T3) between the groups (W=99; P=.36; r=0.07). None of the outcomes showed significant change in accordance with our hypotheses. The intervention group completed a median of 61% (55/90; range 2%-99%) of the exercises in the predefined training program. Adherence was higher in the first intervention period with a specified schedule. CONCLUSIONS: The intervention had no significant effect on postural balance or other outcomes, but the wide range of adherence and a limited sample size challenged the robustness of these conclusions. Future work should increase focus on improving adherence to digital interventions. TRIAL REGISTRATION: ClinicalTrials.gov NCT04364243; https://clinicaltrials.gov/ct2/show/NCT04364243. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.2196/26982.

Multi-Sensor Based IoT Indoor Localization

A typical indoor localization system relies on the availability of infrastructure such as Wi-Fi Access Points, blue-tooth beacons or antenna arrays. This increases the overall system cost and it may not be feasible for deployment in real environments such as shopping malls. A practical indoor localization system should be one that can function with mini-mum existing infrastructure. The proposed system in this paper leverages on the embedded sensors in off-the-shelf Internet of Things (IoT) devices such as smartphone in conjunction with Quick Response (QR) codes which are widely deployed under the authorities requirement due to COVID-19 pandemic. Our proposed stationary inertial measurement unit (IMU) feature is implemented through a first order finite impulse response (FIR) filter that works along with the QR codes. It has successfully reduced the drift errors suffered by IMU. The performance was evaluated in the testing environment at an university campus. From the evaluation results, the proposed method outperformed the conventional method (IMU only) and hybrid model (IMU + QR code) by 94.9% and 57.7% respectively, making the proposed method a promising technique that can be readily applied to other indoor environments. © 2021 IEEE.

A Low-Cost Telerehabilitation Paradigm for Bimanual Training

The COVID-19 pandemic has transformed daily life, as individuals must reduce contacts among each other to prevent the spread of the disease. Consequently, patients’ access to outpatient rehabilitation care was curtailed and their prospect for recovery has been compromised. Telerehabilitation has the potential to provide these patients with equally efficacious therapy in their homes. Using commercial gaming devices with embedded motion sensors, data on movement can be collected toward objective assessment of motor performance, followed by training and documentation of progress. Herein, we present a low-cost telerehabilitation system dedicated to bimanual exercise, wherein the healthy arm drives movements of the affected arm. In the proposed setting, a patient manipulates a dowel embedded with a sensor in front of a Microsoft Kinect sensor. In order to provide an engaging environment for the exercise, the dowel is interfaced with a personal computer, to serve as a controller. The patient’s gestures are translated into actions in a custom-made citizen-science project. Along with the system, we introduce an algorithm for classification of the bimanual movements, whose inner workings are detailed in terms of the procedures performed for dimensionality reduction, feature extraction, and movement classification. We demonstrate the feasibility of our system on eight healthy subjects, offering support to the validity of the algorithm. These preliminary findings set forth the development of precise motion analysis algorithms in affordable home-based rehabilitation.

Validity of a Novel Research-Grade Physical Activity and Sleep Monitor for Continuous Remote Patient Monitoring.

In the midst of the COVID-19 pandemic, Remote Patient Monitoring technologies are highly important for clinicians and researchers. These connected-health technologies enable monitoring of patients and facilitate remote clinical trial research while reducing the potential for the spread of the novel coronavirus. There is a growing requirement for monitoring of the full 24 h spectrum of behaviours with a single research-grade sensor. This research describes a free-living and supervised protocol comparison study of the Verisense inertial measurement unit to assess physical activity and sleep parameters and compares it with the Actiwatch 2 actigraph. Fifteen adults (11 males, 23.4 ± 3.4 years and 4 females, 29 ± 12.6 years) wore both monitors for 2 consecutive days and nights in the free-living study while twelve adults (11 males, 23.4 ± 3.4 years and 1 female, 22 ± 0 years) wore both monitors for the duration of a gym-based supervised protocol study. Agreement of physical activity epoch-by-epoch data with activity classification of sedentary, light and moderate-to-vigorous activity and sleep metrics were evaluated using Spearman's rank-order correlation coefficients and Bland-Altman plots. For all activity, Verisense showed high agreement for both free-living and supervised protocol of r = 0.85 and r = 0.78, respectively. For physical activity classification, Verisense showed high agreement of sedentary activity of r = 0.72 for free-living but low agreement of r = 0.36 for supervised protocol; low agreement of light activity of r = 0.42 for free-living and negligible agreement of r = -0.04 for supervised protocol; and moderate agreement of moderate-to-vigorous activity of r = 0.52 for free-living with low agreement of r = 0.49 for supervised protocol. For sleep metrics, Verisense showed moderate agreement for sleep time and total sleep time of r = 0.66 and 0.54, respectively, but demonstrated high agreement for determination of wake time of r = 0.83. Overall, our results showed moderate-high agreement of Verisense with Actiwatch 2 for assessing epoch-by-epoch physical activity and sleep, but a lack of agreement for activity classifications. Future validation work of Verisense for activity cut-point potentially holds promise for 24 h continuous remote patient monitoring.