Over the past century, in most countries, there has been a continual increase in life expectancy primarily due to improvements in public health, nutrition, personal hygiene and medicine. However, these improvements are now coupled with aging population demographics and falling birth rates, which, when combined, are expected to significantly burden the socioeconomic well-being of many of these countries. In fact, never before in human history have we been confronted with such a large aging population, nor have we developed solid, cost-effective solutions for the well-being, healthcare and social needs of the elderly. One efficient and cost-effective solution to the problem of elderly/patient care is remote healthcare monitoring so they can continue to live at home rather than in nursing homes or hospitals that are very expensive and with limited spaces. These remote monitoring systems will allow medical personnel to keep track of important physiological signs with reduced human resources, at less cost and in real time. This paper introduces several low-cost, noninvasive, user-friendly sensing and actuating systems using information and communication technologies. Such systems can be used to create engineering solutions to some of the pressing healthcare problems in our society, especially as it pertains to the elderly. One example is the integration of sensors, wireless communications, low-power electronics and intelligent computing to determine health-related information using signals from walking patterns. Such a sensing system will be suitable for prolonged use in a home environment. It will be wearable, noninvasive and non-intrusive, similar to smart socks, smart wrist-bands or smart belts. Other examples such as a smart joint monitor and a smart sleeping environment will be discussed, and future perspectives and research challenges in smart home technologies will be described.

Information and communications technologies for elderly ubiquitous healthcare in a smart home

Analysis of lower limb exercises is traditionally completed with four distinct methods: (1) 3D motion capture; (2) depth-camera-based systems; (3) visual analysis from a qualified exercise professional; and (4) self-assessment. Each method is associated with a number of limitations. The aim of this systematic review is to synthesise and evaluate studies which have investigated the capacity for inertial measurement unit (IMU) technologies to assess movement quality in lower limb exercises. A systematic review of studies identified through the databases of PubMed, ScienceDirect and Scopus was conducted. Articles written in English and published in the last 10 years which investigated an IMU system for the analysis of repetition-based targeted lower limb exercises were included. The quality of included studies was measured using an adapted version of the STROBE assessment criteria for cross-sectional studies. The studies were categorised into three groupings: exercise detection, movement classification or measurement validation. Each study was then qualitatively summarised. From the 2452 articles that were identified with the search strategies, 47 papers are included in this review. Twenty-six of the 47 included studies were deemed as being of high quality. Wearable inertial sensor systems for analysing lower limb exercises is a rapidly growing field of research. Research over the past 10 years has predominantly focused on validating measurements that the systems produce and classifying users’ exercise quality. There have been very few user evaluation studies and no clinical trials in this field to date.

/pdf/wearable-inertial-sensor-systems-for-lower-limb-exercise-4nkwiib87b.pdf

Wearable Inertial Sensor Systems for Lower Limb Exercise Detection and Evaluation: A Systematic Review

Computer-aided assessment of physical rehabilitation entails evaluation of patient performance in completing prescribed rehabilitation exercises, based on processing movement data captured with a sensory system. Despite the essential role of rehabilitation assessment toward improved patient outcomes and reduced healthcare costs, existing approaches lack versatility, robustness, and practical relevance. In this paper, we propose a deep learning-based framework for automated assessment of the quality of physical rehabilitation exercises. The main components of the framework are metrics for quantifying movement performance, scoring functions for mapping the performance metrics into numerical scores of movement quality, and deep neural network models for generating quality scores of input movements via supervised learning. The proposed performance metric is defined based on the log-likelihood of a Gaussian mixture model, and encodes low-dimensional data representation obtained with a deep autoencoder network. The proposed deep spatio-temporal neural network arranges data into temporal pyramids, and exploits the spatial characteristics of human movements by using sub-networks to process joint displacements of individual body parts. The presented framework is validated using a dataset of ten rehabilitation exercises. The significance of this work is that it is the first that implements deep neural networks for assessment of rehabilitation performance.

/pdf/a-deep-learning-framework-for-assessing-physical-2bwd961n4c.pdf

A Deep Learning Framework for Assessing Physical Rehabilitation Exercises

Accurate assessments of adherence and exercise performance are required in order to ensure that patients adhere to and perform their rehabilitation exercises correctly within the home environment. Inertial sensors have previously been advocated as a means of achieving these requirements, by using them as an input to an exercise biofeedback system. This research sought to investigate whether inertial sensors, and in particular a single sensor, can accurately classify exercise performance in patients performing lower limb exercises for rehabilitation purposes. Fifty-eight participants (19 male, 39 female, age: 53.9 ± 8.5 years, height: 1.69 ± 0.08 m, weight: 74.3 ± 13.0 kg) performed ten repetitions of seven lower limb exercises (hip abduction, hip flexion, hip extension, knee extension, heel slide, straight leg raise, and inner range quadriceps). Three inertial sensor units, secured to the thigh, shin and foot of the leg being exercised, were used to acquire data during each exercise. Machine learning classification methods were applied to quantify the acquired data. The classification methods achieved relatively high accuracy at distinguishing between correct and incorrect performance of an exercise using three, two, or one sensor while moderate efficacy scores were also achieved by the classifier when attempting to classify the particular error in exercise performance. Results also illustrated that a reduction in the number of inertial sensor units employed has little effect on the overall efficacy results. The results revealed that it is possible to classify lower limb exercise performance using inertial sensors with satisfactory levels of accuracy and reducing the number of sensors employed does not reduce the accuracy of the method.

/pdf/rehabilitation-exercise-assessment-using-inertial-sensors-a-th2ltze0xe.pdf

Rehabilitation exercise assessment using inertial sensors: a cross-sectional analytical study

Since the knee joint bears the full weight load of the human body and the highest pressure loads while providing flexible movement, it is the body part most vulnerable and susceptible to osteoarthritis. In exercise therapy, the early rehabilitation stages last for approximately six weeks, during which the patient works with the physical therapist several times each week. The patient is afterwards given instructions for continuing rehabilitation exercise by him/herself at home. This study develops a rehabilitation exercise assessment mechanism using three wearable sensors mounted on the chest, thigh and shank of the working leg in order to enable the patients with knee osteoarthritis to manage their own rehabilitation progress. In this work, time-domain, frequency-domain features and angle information of the motion sensor signals are used to classify the exercise type and identify whether their postures are proper or not. Three types of rehabilitation exercise commonly prescribed to knee osteoarthritis patients are: Short-Arc Exercise, Straight Leg Raise, and Quadriceps Strengthening Mini-squats. After ten subjects performed the three kinds of rehabilitation activities, three validation techniques including 10-fold cross-validation, within subject cross validation, and leave-one-subject cross validation are utilized to confirm the proposed mechanism. The overall recognition accuracy for exercise type classification is 97.29% and for exercise posture identification it is 88.26%. The experimental results demonstrate the feasibility of the proposed mechanism which can help patients perform rehabilitation movements and progress effectively. Moreover, the proposed mechanism is able to detect multiple errors at once, fulfilling the requirements for rehabilitation assessment.

/pdf/wearable-sensor-based-rehabilitation-exercise-assessment-for-17p5hhgoh4.pdf

Wearable sensor-based rehabilitation exercise assessment for knee osteoarthritis

In this paper, we describe methods for assessment of exercise quality using body-worn tri-axial accelerometers. We assess exercise quality by building a classifier that labels incorrect exercises. The incorrect performances are divided into a number of classes of errors as defined by a physical therapist. We focus on exercises commonly prescribed for knee osteoarthritis: standing hamstring curl, reverse hip abduction, and lying straight leg raise. The methods presented here will form the basis for an at-home rehabilitation device that will recognize errors in patient exercise performance, provide appropriate feedback on the performance, and motivate the patient to continue the prescribed regimen.

Classifying human motion quality for knee osteoarthritis using accelerometers

Knowing how well an activity is performed is important for home rehabilitation. We would like to not only know if a motion is being performed correctly, but also in what way the motion is incorrect so that we may provide feedback to the user. This paper describes methods for assessing human motion quality using body-worn tri-axial accelerometers and gyroscopes. We use multi-label classifiers to detect subtle errors in exercise performances of eight individuals with knee osteoarthritis, a degenerative disease of the cartilage. We present results obtained using various machine learning methods with decision tree base classifiers. The classifier can detect classes in multi-label data with 75% sensitivity, 90% specificity and 80% accuracy. The methods presented here form the basis for an at-home rehabilitation device that will recognize errors in patient exercise performance, provide appropriate feedback on the performance, and motivate the patient to continue the prescribed regimen.

Portia E. Taylor

Papers

Classifying human motion quality for knee osteoarthritis using accelerometers

Multi-label classification for the analysis of human motion quality