Limb Movements Classification Using Wearable Wireless Transceivers

Abstract

Abstract-A feasibility study where small wireless transceivers are used to classify some typical limb movements used in physical therapy processes is presented. Wearable wireless low-cost commercial transceivers operating at 2.4 GHz are supposed to be widely deployed in indoor settings and on people's bodies in tomorrow's pervasive computing environments. The key idea of this work is to exploit their presence by collecting the received signal strength measured between those worn by a person. The measurements are used to classify a set of kinesiotherapy activities. The collected data are classified using both Support Vector Machine and K-Nearest Neighbour methods, in order to recognise the different activities. Index Terms- I. INTRODUCTIONHE increasing improvements of wireless technology and miniaturized sensors made possible the proliferation of wireless sensor networks. These communications networks are one of the first real world examples of pervasive computing: they are composed of small, smart and cheap sensing devices that promise to eventually permeate the environment. We may face a not so distant future where small sensors, all capable of wireless communication, are ubiquitously deployed in the environment and on people's body. TExtensive attention has been focused in the literature on wireless sensor networks, especially in the framework of wireless body sensor networks [1]. In particular, wearable wireless systems were developed to detect, track and understand people's behaviour. In recent years, the detection of body posture and activity received a significant interest for their application in sports, medicine and military. Recognizing people's activities is also a key issue in Assisted Living (AL) applications, where it can be useful for example to rate how a person performs routine activities [2,3].One further field of application is kinesiotherapy, where the aim is to provide monitoring of physical therapies for patients who have suffered a stroke, multiple sclerosis, joint replacements or reconstructions, amputation, brain and spinal cord injury, or some motor function disability resulting from Parkinson's disease [4][5][6]. For these cases, wireless body  Manuscript received….Anda R. Guraliuc and Paolo Nepa are with the Department of Information Engineering, University of Pisa, Italy (via Caruso 16, I-56122, Pisa, Italy, phone: +39 050 2217511; fax: +39 050 2217522; e-mail: {anda.guraliuc, paolo.nepa}@iet.unipi.it).Paolo Barsocchi and Francesco Potortì are with the ISTI Institute of CNR (via Moruzzi 1, I-56124, Pisa, Italy, e-mail: {paolo.barsocchi, potorti}@isti.cnr.it): their work is supported in part by the European Commission in the framework of the FP7 project UNIVERSAAL (contract N. 247950).sensor networks could replace the existing wired telemetry systems [7], allowing remotely supervised kinesiotherapy. In a typical wearable Wireless Body Area Network (WBAN) scenario, a patient wears some sensors that form an on-body sensor network, while an off-body base station registers data collected...

Movement Detection of Human Body Segments: Passive radio-frequency identification and machine-learning technologies

“…In simpler systems with no dedicated sensors [6], motion information is gathered from the electromagnetic signals measured between wireless active transceiver modules placed on the body and an off-body fixed node forming a wireless network. The analysis of the fluctuations of the received signal strength (RSS) due to relative motion of body segments is performed by classification algorithms with the purpose of recognizing specific periodic motor patterns.…”

“…The purpose of this experiment is to compare the classification capability of batteryless wearable tags with those recently demonstrated for active systems involving true transmitters [6]. The classification of periodic motor patterns concerns the recognition of one of eight predefined kinesiotherapy activities involving limb motions that are typical of a rehabilitation/recovery process [see Figure 3 …”

“…Subjects are asked to perform three cyclic movements involving the arms and five involving the legs (see the insets in Figure 7) as in [6]. During the measurements, the activities are performed at a controlled rate of 30 periodic movements/min.…”

Impulse Radio Ultra-Wideband Communications for Localization and Tracking of Human Body and Limbs Movement for Healthcare Applications

“…In particular, wearable wireless systems are attracting significant interest to access and monitor human activity for sports, healthcare, military applications and day-to-day activities [1][2][3][4]. These include various technologies such as as infrared, inertial, ultrasound, optical and Radio Frequency (RF) based systems [4][5].…”

Physiological Features from an On-Body Radio Propagation Channel

“…In [7,8], the received signal strength (RSS), an internal parameter of RF transceivers, was used to characterize limb movements for kinesio-therapy activities using Crossbow IRIS wireless modules. Their results show that learning techniques, such as a Support Vector Machine (SVM), can determine and classify limb movements based on information from the fading in the propagation signal path.…”