**5. Lessons learned through the research, test, and evaluation of the Wireless Quantified Reflex Device for the broader evolution of Network Centric Therapy, such as gait and movement disorder quantification**

A readily noted capability observed by LeMoyne and Mastroianni was that since the wireless accelerometer was functionally wearable for the quantification of reflex response through mounting about the lateral malleolus of the patellar

**297**

*An Evolutionary Perspective for Network Centric Therapy through Wearable and Wireless…*

tendon, likewise the same mounting procedure could be applied for quantifying gait patterns [11, 12, 17]. Alternative mounting configurations, such as the lateral epicondyle proximal to the knee, were also feasible for assessing gait in a quantified context [11]. Additionally, the smartphone and portable media device were suitable candidates to represent functionally wearable and wireless inertial sensor systems (both accelerometers and gyroscopes) for gait quantification and eventually machine learning classification [3–10]. These concepts were also applied to the quantification of movement disorders through the mounting of wearable and

wireless inertial sensor systems about the dorsum of the hand [1, 2, 5–10].

**6. Evolutionary pathway for Network Centric Therapy with respect to** 

Preliminary attempts to apply functionally wearable wireless accelerometers to measure gait characteristics consisted of segmented subsystems and in some cases complex mounting techniques exceeding the knowledge of the common user [69–72]. The highly miniaturized, portable, and non-intrusive nature of the G-link wireless accelerometer developed by Microstrain demonstrate its robust capability for quantifying gait characteristics [11]. Proof of concept from an engineering perspective was demonstrated for the identification of quantified disparity of hemiplegic gait and Virtual Proprioception to enable real-time rehabilitation of hemiplegic gait [73–76]. Preliminary research, development, testing, and evaluation by LeMoyne et al. applied the G-link Microstrain wireless accelerometer to ascertain quantified disparity of hemiplegic gait. The wireless accelerometer nodes were effectively wearable. They could be mounted about the lateral epicondyle proximal to the knee through an elastic band or about the lateral malleolus near the ankle using the

The wireless accelerometer achieved connectivity to a locally situated personal computer, which would then serve as the basis for post-processing. Using the acceleration magnitude of the three-dimensional orthogonal acceleration signal, characteristic spikes of the acceleration magnitude signal represented the initiation of stance. The time averaged acceleration from stance to stance enabled the quantification of gait characteristics [74, 75]. Furthermore, through the ratio of the hemiplegic affected leg to the unaffected leg using the time averaged acceleration from stance to stance, the quantified disparity of hemiplegic gait could be quantified with the potential for deriving therapeutic intervention for rehabilitation [74]. Functionally wearable and locally wireless accelerometers have also been applied to successfully contrast hemiplegic gait with respect to the frequency domain [73]. Other applications of wireless accelerometer systems that are functionally wearable have been successfully demonstrated for the context of effectively autonomous gait analysis based on quantified data derived from the acceleration signal [11, 12, 69–72]. Virtual Proprioception expanded the capabilities of functionally wearable wireless accelerometers for real-time modification of hemiplegic gait based on accelerometer signal data. The wireless accelerometers were mounted by flexible elastic bands proximal to the lateral epicondyle of the knee for both the unaffected leg and hemiplegic affected leg. Based on a visual feedback strategy the person with hemiplegic gait was able to modify the hemiplegic affected leg to a more representa-

tive acceleration signal representative of the unaffected leg [76].

During 2010 LeMoyne and Mastroianni sought to expand the availability of wearable and wireless accelerometer systems for quantifying gait in the context of more commercially available systems. The smartphone of that timeframe was equipped with an internal accelerometer. Additionally, the smartphone is inherently capable of wirelessly accessing the Internet. A software application for recording

*DOI: http://dx.doi.org/10.5772/intechopen.95550*

**quantification of gait**

elastic band of a sock [74, 75].

*An Evolutionary Perspective for Network Centric Therapy through Wearable and Wireless… DOI: http://dx.doi.org/10.5772/intechopen.95550*

tendon, likewise the same mounting procedure could be applied for quantifying gait patterns [11, 12, 17]. Alternative mounting configurations, such as the lateral epicondyle proximal to the knee, were also feasible for assessing gait in a quantified context [11]. Additionally, the smartphone and portable media device were suitable candidates to represent functionally wearable and wireless inertial sensor systems (both accelerometers and gyroscopes) for gait quantification and eventually machine learning classification [3–10]. These concepts were also applied to the quantification of movement disorders through the mounting of wearable and wireless inertial sensor systems about the dorsum of the hand [1, 2, 5–10].

### **6. Evolutionary pathway for Network Centric Therapy with respect to quantification of gait**

Preliminary attempts to apply functionally wearable wireless accelerometers to measure gait characteristics consisted of segmented subsystems and in some cases complex mounting techniques exceeding the knowledge of the common user [69–72]. The highly miniaturized, portable, and non-intrusive nature of the G-link wireless accelerometer developed by Microstrain demonstrate its robust capability for quantifying gait characteristics [11]. Proof of concept from an engineering perspective was demonstrated for the identification of quantified disparity of hemiplegic gait and Virtual Proprioception to enable real-time rehabilitation of hemiplegic gait [73–76].

Preliminary research, development, testing, and evaluation by LeMoyne et al. applied the G-link Microstrain wireless accelerometer to ascertain quantified disparity of hemiplegic gait. The wireless accelerometer nodes were effectively wearable. They could be mounted about the lateral epicondyle proximal to the knee through an elastic band or about the lateral malleolus near the ankle using the elastic band of a sock [74, 75].

The wireless accelerometer achieved connectivity to a locally situated personal computer, which would then serve as the basis for post-processing. Using the acceleration magnitude of the three-dimensional orthogonal acceleration signal, characteristic spikes of the acceleration magnitude signal represented the initiation of stance. The time averaged acceleration from stance to stance enabled the quantification of gait characteristics [74, 75]. Furthermore, through the ratio of the hemiplegic affected leg to the unaffected leg using the time averaged acceleration from stance to stance, the quantified disparity of hemiplegic gait could be quantified with the potential for deriving therapeutic intervention for rehabilitation [74]. Functionally wearable and locally wireless accelerometers have also been applied to successfully contrast hemiplegic gait with respect to the frequency domain [73]. Other applications of wireless accelerometer systems that are functionally wearable have been successfully demonstrated for the context of effectively autonomous gait analysis based on quantified data derived from the acceleration signal [11, 12, 69–72].

Virtual Proprioception expanded the capabilities of functionally wearable wireless accelerometers for real-time modification of hemiplegic gait based on accelerometer signal data. The wireless accelerometers were mounted by flexible elastic bands proximal to the lateral epicondyle of the knee for both the unaffected leg and hemiplegic affected leg. Based on a visual feedback strategy the person with hemiplegic gait was able to modify the hemiplegic affected leg to a more representative acceleration signal representative of the unaffected leg [76].

During 2010 LeMoyne and Mastroianni sought to expand the availability of wearable and wireless accelerometer systems for quantifying gait in the context of more commercially available systems. The smartphone of that timeframe was equipped with an internal accelerometer. Additionally, the smartphone is inherently capable of wirelessly accessing the Internet. A software application for recording

*Wireless Sensor Networks - Design, Deployment and Applications*

**5. Lessons learned through the research, test, and evaluation of the Wireless Quantified Reflex Device for the broader evolution of Network Centric Therapy, such as gait and movement disorder** 

*The J48 decision tree to distinguish between a hemiplegic affected leg and unaffected leg, for which the time disparity between maximum and minimum angular rate of rotation for the patellar tendon reflex response numeric attribute is illustrated as the most prevalent for establishing classification accuracy [67].*

*The gyroscope signal of the patellar tendon reflex response for the unaffected leg using the potential energy* 

A readily noted capability observed by LeMoyne and Mastroianni was that since the wireless accelerometer was functionally wearable for the quantification of reflex response through mounting about the lateral malleolus of the patellar

**296**

**Figure 3.**

**Figure 2.**

*impact pendulum to evoke the reflex [67].*

**quantification**

the accelerometer data for a prescribed duration and sampling rate with wireless transfer to the Internet as an email attachment enables the smartphone to function as a wearable and wireless inertial sensor system. The email resource represents a provisional representation of a Cloud computing resource. These characteristics enable the smartphone to quantify gait features in the context of a wearable and wireless inertial sensor system [77]. These preliminary capabilities constitute the origins of Network Centric Therapy for the domain of gait analysis [3–7, 78].

Preliminary testing and evaluation of the smartphone as a wearable and wireless inertial sensor system for gait analysis was conducted in region of Pittsburgh, Pennsylvania. The experimental gait analysis accelerometer data was conveyed wirelessly to the Internet as an email attachment for subsequent post-processing in the general area of Los Angeles, California. The implications were that experimental and post-processing locations could be geographically separated anywhere in the world with Internet connectivity [77].

The preliminary gait experiment of 2010 implementing the smartphone as a wearable and wireless inertial sensor system through the internal accelerometer involved mounting the smartphone proximal to the lateral malleolus of the ankle joint by an elastic band. Two primary gait characteristics were quantified, such as the temporal duration between stance to stance and time averaged acceleration from stance to stance. These parameters acquired by the smartphone functioning as a wearable and wireless inertial sensor system through the available accelerometer demonstrated considerable accuracy and reliability [77].

Additional and similar themed experiments pertained to quantification of gait through other mounting applications, which underscores the flexibility of the smartphone as a wearable and wireless inertial sensor system. The two other mounting positions involved the lateral epicondyle near the knee joint and lumbarsacral aspect of the spine through an elastic band. The temporal duration between stance to stance cycle displayed considerable accuracy and reliability and successfully elucidated predominant frequencies in the context of the frequency domain with respect to both mounting strategies [79, 80].

Another device that is similar to the smartphone for applications as a wearable and wireless inertial sensor system for the quantification of gait is the portable media device. The portable media device can utilize the same software application as relevant to the smartphone. Although the portable media device is generally restricted to an area with local internet connectivity, this device has a lighter mass and is more affordable for tandem applications involving both legs for gait analysis [3–10, 78].

Preliminary, testing of the portable media device was successfully demonstrated with mounting about the lateral malleolus of the leg by an elastic band. The accelerometer of the portable media device successfully quantified gait in an accurate and consistent manner. The experimental data was conveyed by wireless transmission through the Internet as an email attachment, and the experimental and post-processing resources were situated on opposite sides of the continental United States. Post-processing emphasized the derivation of step cycle time (stance to stance) and time averaged acceleration (stance to stance) [81].

An observation of the portable media device is that it is more affordable than the smartphone, such as for the application of two tandem operated portable media devices for quantifying the disparity of hemiplegic gait. LeMoyne and Mastroianni incorporated two portable media devices in the context of a wearable and wireless inertial sensor system, such as an accelerometer, for quantifying hemiplegic gait respective of the unaffected leg and the hemiplegic affected leg. The devices were mounted about the lateral malleolus of the ankle joint through an elastic band for both the unaffected leg and the hemiplegic affected leg. The tandem activated portable media devices successfully demonstrated the ability to quantitatively identify

**299**

*An Evolutionary Perspective for Network Centric Therapy through Wearable and Wireless…*

stance to stance temporal duration and stance to stance time averaged acceleration of the hemiplegic affected leg and unaffected leg with statistical significance. Also, the ratio of stance to stance time averaged acceleration less the offset for the hemiplegic affected leg and unaffected leg demonstrated quantified disparity [82]. Eventually a strategy for using a singular smartphone to quantify hemiplegic gait and its associated disparity was established with the incorporation of a treadmill to maintain constant velocity. The smartphone functioning as a wearable and wireless accelerometer platform was mounted about the lateral malleolus of the ankle by an elastic band. Automated post-processing software emphasized the rhythmic characteristics of gait and acquired gait parameters, such as stance to stance temporal disparity and stance to stance time averaged acceleration. The stance to stance temporal disparity did not display statistical significance, because of the treadmill velocity constraint. Statistical significance was achieved for stance to stance time averaged acceleration with respect to comparing the hemiplegic affected leg to the unaffected leg. This experimental configuration enables the evaluation and quanti-

Evolutionary trends eventually enabled the smartphone to quantify gait through the internal gyroscope, which offers a more clinically representative kinematic signal. The strategy of conducting gait analysis constrained to a constant velocity by a treadmill was applied. A smartphone functioning as a wearable and wireless gyroscope platform quantified hemiplegic gait in terms of both the affected leg and unaffected leg with mounting about the lateral malleolus near the ankle joint through an elastic band. The gyroscope signal was consolidated to a feature set during the post-processing phase, which consisted of five numeric attributes: maximum, minimum, mean, standard deviation, and coefficient of variation. Using the multilayer perceptron neural network considerable classification accuracy was attained for distinguishing

between the hemiplegic affected leg and unaffected leg during gait [84].

Additionally, the smartphone through its internal inertial sensor system has been applied to other applications pertaining to the domain of gait analysis and associated mobility. Smartphones have been successfully incorporated for augmenting the acuity of clinically standard evaluations, such as the Timed Up and Go and 6-Minute Walk Test [85–87]. An observed utility of the strategy of augmenting clinically standard evaluation techniques with functionally wearable and wireless inertial sensor systems, such as the smartphone, is the ability to evolve a clinical

During this phase of the evolutionary process that lead to the realization of Network Centric Therapy a new observation occurred. Smartphones and portable media devices can function as representative and effective wearable and wireless inertial sensor systems. However, their evolutionary pathway is not consistent with the biomedical and healthcare domain. A new perspective for wearable and wireless inertial sensor systems was developed, which incorporated inertial sensor nodes with local wireless connectivity to a device, such as a smartphone or tablet, with considerably expanded wireless access to the Internet. This paradigm shift enabled considerable reduction in mass and volumetric profile for the wearable and wireless inertial sensor system. This strategy enabled segmented wireless access of the

inertial signal data for connectivity to a Cloud computing resource [88].

During 2016 LeMoyne et al. utilized a wearable and wireless inertial sensor system architecture in the context of Network Centric Therapy for the evaluation of gait for subject's with Friedreich's ataxia. The system applied local wearable and wireless inertial sensor nodes with local wireless connectivity to a tablet with global wireless access to a Cloud computing environment. A multilayer perceptron neural network achieved considerable classification accuracy to distinguish between a person with healthy gait and gait for a person with Friedreich's ataxia [89].

*DOI: http://dx.doi.org/10.5772/intechopen.95550*

fication of gait in an autonomous environment [83].

method rather than inventing a new methodology.

#### *An Evolutionary Perspective for Network Centric Therapy through Wearable and Wireless… DOI: http://dx.doi.org/10.5772/intechopen.95550*

stance to stance temporal duration and stance to stance time averaged acceleration of the hemiplegic affected leg and unaffected leg with statistical significance. Also, the ratio of stance to stance time averaged acceleration less the offset for the hemiplegic affected leg and unaffected leg demonstrated quantified disparity [82].

Eventually a strategy for using a singular smartphone to quantify hemiplegic gait and its associated disparity was established with the incorporation of a treadmill to maintain constant velocity. The smartphone functioning as a wearable and wireless accelerometer platform was mounted about the lateral malleolus of the ankle by an elastic band. Automated post-processing software emphasized the rhythmic characteristics of gait and acquired gait parameters, such as stance to stance temporal disparity and stance to stance time averaged acceleration. The stance to stance temporal disparity did not display statistical significance, because of the treadmill velocity constraint. Statistical significance was achieved for stance to stance time averaged acceleration with respect to comparing the hemiplegic affected leg to the unaffected leg. This experimental configuration enables the evaluation and quantification of gait in an autonomous environment [83].

Evolutionary trends eventually enabled the smartphone to quantify gait through the internal gyroscope, which offers a more clinically representative kinematic signal. The strategy of conducting gait analysis constrained to a constant velocity by a treadmill was applied. A smartphone functioning as a wearable and wireless gyroscope platform quantified hemiplegic gait in terms of both the affected leg and unaffected leg with mounting about the lateral malleolus near the ankle joint through an elastic band. The gyroscope signal was consolidated to a feature set during the post-processing phase, which consisted of five numeric attributes: maximum, minimum, mean, standard deviation, and coefficient of variation. Using the multilayer perceptron neural network considerable classification accuracy was attained for distinguishing between the hemiplegic affected leg and unaffected leg during gait [84].

Additionally, the smartphone through its internal inertial sensor system has been applied to other applications pertaining to the domain of gait analysis and associated mobility. Smartphones have been successfully incorporated for augmenting the acuity of clinically standard evaluations, such as the Timed Up and Go and 6-Minute Walk Test [85–87]. An observed utility of the strategy of augmenting clinically standard evaluation techniques with functionally wearable and wireless inertial sensor systems, such as the smartphone, is the ability to evolve a clinical method rather than inventing a new methodology.

During this phase of the evolutionary process that lead to the realization of Network Centric Therapy a new observation occurred. Smartphones and portable media devices can function as representative and effective wearable and wireless inertial sensor systems. However, their evolutionary pathway is not consistent with the biomedical and healthcare domain. A new perspective for wearable and wireless inertial sensor systems was developed, which incorporated inertial sensor nodes with local wireless connectivity to a device, such as a smartphone or tablet, with considerably expanded wireless access to the Internet. This paradigm shift enabled considerable reduction in mass and volumetric profile for the wearable and wireless inertial sensor system. This strategy enabled segmented wireless access of the inertial signal data for connectivity to a Cloud computing resource [88].

During 2016 LeMoyne et al. utilized a wearable and wireless inertial sensor system architecture in the context of Network Centric Therapy for the evaluation of gait for subject's with Friedreich's ataxia. The system applied local wearable and wireless inertial sensor nodes with local wireless connectivity to a tablet with global wireless access to a Cloud computing environment. A multilayer perceptron neural network achieved considerable classification accuracy to distinguish between a person with healthy gait and gait for a person with Friedreich's ataxia [89].

*Wireless Sensor Networks - Design, Deployment and Applications*

world with Internet connectivity [77].

demonstrated considerable accuracy and reliability [77].

with respect to both mounting strategies [79, 80].

time averaged acceleration (stance to stance) [81].

the accelerometer data for a prescribed duration and sampling rate with wireless transfer to the Internet as an email attachment enables the smartphone to function as a wearable and wireless inertial sensor system. The email resource represents a provisional representation of a Cloud computing resource. These characteristics enable the smartphone to quantify gait features in the context of a wearable and wireless inertial sensor system [77]. These preliminary capabilities constitute the origins of Network Centric Therapy for the domain of gait analysis [3–7, 78].

Preliminary testing and evaluation of the smartphone as a wearable and wireless inertial sensor system for gait analysis was conducted in region of Pittsburgh, Pennsylvania. The experimental gait analysis accelerometer data was conveyed wirelessly to the Internet as an email attachment for subsequent post-processing in the general area of Los Angeles, California. The implications were that experimental and post-processing locations could be geographically separated anywhere in the

The preliminary gait experiment of 2010 implementing the smartphone as a wearable and wireless inertial sensor system through the internal accelerometer involved mounting the smartphone proximal to the lateral malleolus of the ankle joint by an elastic band. Two primary gait characteristics were quantified, such as the temporal duration between stance to stance and time averaged acceleration from stance to stance. These parameters acquired by the smartphone functioning as a wearable and wireless inertial sensor system through the available accelerometer

Additional and similar themed experiments pertained to quantification of gait through other mounting applications, which underscores the flexibility of the smartphone as a wearable and wireless inertial sensor system. The two other mounting positions involved the lateral epicondyle near the knee joint and lumbarsacral aspect of the spine through an elastic band. The temporal duration between stance to stance cycle displayed considerable accuracy and reliability and successfully elucidated predominant frequencies in the context of the frequency domain

Another device that is similar to the smartphone for applications as a wearable and wireless inertial sensor system for the quantification of gait is the portable media device. The portable media device can utilize the same software application as relevant to the smartphone. Although the portable media device is generally restricted to an area with local internet connectivity, this device has a lighter mass and is more affordable for tandem applications involving both legs for gait analysis [3–10, 78]. Preliminary, testing of the portable media device was successfully demonstrated with mounting about the lateral malleolus of the leg by an elastic band. The accelerometer of the portable media device successfully quantified gait in an accurate and consistent manner. The experimental data was conveyed by wireless transmission through the Internet as an email attachment, and the experimental and post-processing resources were situated on opposite sides of the continental United States. Post-processing emphasized the derivation of step cycle time (stance to stance) and

An observation of the portable media device is that it is more affordable than the smartphone, such as for the application of two tandem operated portable media devices for quantifying the disparity of hemiplegic gait. LeMoyne and Mastroianni incorporated two portable media devices in the context of a wearable and wireless inertial sensor system, such as an accelerometer, for quantifying hemiplegic gait respective of the unaffected leg and the hemiplegic affected leg. The devices were mounted about the lateral malleolus of the ankle joint through an elastic band for both the unaffected leg and the hemiplegic affected leg. The tandem activated portable media devices successfully demonstrated the ability to quantitatively identify

**298**

The current state of the art for demonstrating the capability of Network Centric Therapy involves the recent test and evaluation of the BioStamp nPoint, which represents a conformal wearable and wireless inertial sensor system. The BioStamp nPoint achieves wireless connectivity for acquiring signal data for quantifying gait in a segmented manner through wireless systems, such as a tablet for operation and smartphone for Cloud computing access. **Figure 4** presents the supporting apparatus for the BioStamp nPoint conformal wearable and wireless inertial sensor system [90].

Recently, during 2020 LeMoyne and Mastroianni applied the BioStamp nPoint to quantify hemiplegic gait with distinction through machine learning. The BioStamp nPoint conformal wearable and wireless inertial sensor system was mounted by adhesive medium to both the hemiplegic affected leg and unaffected leg about the femur and proximal to the patella as shown in **Figure 5**. The subject walked on a treadmill for the experiment [91].

#### **Figure 4.**

*The BioStamp nPoint conformal wearable and wireless inertial sensor system and supporting devices, such as docking station, tablet, and smartphone [90].*

#### **Figure 5.**

*The BioStamp nPoint conformal wearable and wireless inertial sensor system mounted about the femur for the quantification of hemiplegic gait [91].*

**301**

**Figure 7.**

**Figure 6.**

*An Evolutionary Perspective for Network Centric Therapy through Wearable and Wireless…*

The gyroscope signal revealed notable disparity respective of the affected leg and unaffected leg during gait as presented in **Figures 6** and **7** respectively. Postprocessing of the gyroscope signal data consolidated a feature set consisting of five numeric attributes based on descriptive statistics, such as maximum, minimum, mean, standard deviation, and coefficient of variation. Multiple machine learning classification algorithms, such as the support vector machine and multilayer perceptron neural network, achieved considerable classification accuracy to distinguish between the hemiplegic affected leg and unaffected leg [91, 92].

*The BioStamp nPoint conformal wearable and wireless inertial sensor signal for the unaffected leg [91].*

*The BioStamp nPoint conformal wearable and wireless inertial sensor signal for the hemiplegic affected leg [91].*

*DOI: http://dx.doi.org/10.5772/intechopen.95550*

*An Evolutionary Perspective for Network Centric Therapy through Wearable and Wireless… DOI: http://dx.doi.org/10.5772/intechopen.95550*

**Figure 6.** *The BioStamp nPoint conformal wearable and wireless inertial sensor signal for the hemiplegic affected leg [91].*

**Figure 7.** *The BioStamp nPoint conformal wearable and wireless inertial sensor signal for the unaffected leg [91].*

The gyroscope signal revealed notable disparity respective of the affected leg and unaffected leg during gait as presented in **Figures 6** and **7** respectively. Postprocessing of the gyroscope signal data consolidated a feature set consisting of five numeric attributes based on descriptive statistics, such as maximum, minimum, mean, standard deviation, and coefficient of variation. Multiple machine learning classification algorithms, such as the support vector machine and multilayer perceptron neural network, achieved considerable classification accuracy to distinguish between the hemiplegic affected leg and unaffected leg [91, 92].

*Wireless Sensor Networks - Design, Deployment and Applications*

treadmill for the experiment [91].

*docking station, tablet, and smartphone [90].*

The current state of the art for demonstrating the capability of Network Centric

*The BioStamp nPoint conformal wearable and wireless inertial sensor system and supporting devices, such as* 

*The BioStamp nPoint conformal wearable and wireless inertial sensor system mounted about the femur for the* 

Therapy involves the recent test and evaluation of the BioStamp nPoint, which represents a conformal wearable and wireless inertial sensor system. The BioStamp nPoint achieves wireless connectivity for acquiring signal data for quantifying gait in a segmented manner through wireless systems, such as a tablet for operation and smartphone for Cloud computing access. **Figure 4** presents the supporting apparatus for the BioStamp nPoint conformal wearable and wireless inertial sensor system [90]. Recently, during 2020 LeMoyne and Mastroianni applied the BioStamp nPoint to quantify hemiplegic gait with distinction through machine learning. The BioStamp nPoint conformal wearable and wireless inertial sensor system was mounted by adhesive medium to both the hemiplegic affected leg and unaffected leg about the femur and proximal to the patella as shown in **Figure 5**. The subject walked on a

**300**

**Figure 5.**

*quantification of hemiplegic gait [91].*

**Figure 4.**
