10 views

Original Title: Muniz 2010

Uploaded by seyedmohamad

- A Real-Time Interference Monitoring Technique for GNSS Based on a Twin Support Vector Machine Method.pdf
- IJEST Special Issue Complete
- Mcq Mock
- R Installation
- Iaetsd-jaras-load Reduction in Videos Using Advanced Fast
- Art. Redes Neuronales de Biomasa
- Fully Automatic Facial Action Unit Detection and Temporal Analysis.pdf
- Articulated Pose Estimation With Flexible Mixtures-Of-parts
- A Review on Construction Cost Forecasting Techniques
- 2532511
- svm
- Fath on y 2017 Adversarial
- IRJET-Intelligent Emotion Detection System using Facial Images
- Hand gesture recognition using support vector machine
- AN EFFICIENT HEMORRHAGE DETECTION USING SVM CLASSIFIER
- document
- Super Parsing Sentiment Classification With Review Extraction
- research proposal machine learning
- 2D 3D Face Recognition
- en_Tanagra_KMO_Bartlett.pdf

You are on page 1of 7

Journal of Biomechanics

journal homepage: www.elsevier.com/locate/jbiomech

www.JBiomech.com

and logistic regression for evaluating the effect of subthalamic stimulation in

Parkinson disease on ground reaction force during gait

A.M.S. Muniz a, H. Liu b, K.E. Lyons c, R. Pahwa c, W. Liu b, F.F. Nobre a, J. Nadal a,n

a

Biomedical Engineering Program, Federal University of Rio de Janeiro, COPPE, P.O. Box 68510, 21941-972, Rio de Janeiro, RJ, Brazil

b

Department of Physical Therapy and Rehabilitation Sciences, University of Kansas Medical Center, Kansas City, KS, USA

c

Department of Neurology, University of Kansas Medical Center, Kansas City, KS, USA

a r t i c l e in f o a b s t r a c t

Article history: Deep brain stimulation of the subthalamic nucleus (DBS-STN) is an approved treatment for advanced

Accepted 6 October 2009 Parkinson disease (PD) patients; however, there is a need to further evaluate its effect on gait. This study

compares logistic regression (LR), probabilistic neural network (PNN) and support vector machine

Keywords: (SVM) classiers for discriminating between normal and PD subjects in assessing the effects of DBS-STN

Parkinson disease on ground reaction force (GRF) with and without medication. Gait analysis of 45 subjects (30 normal

Deep brain stimulation and 15 PD subjects who underwent bilateral DBS-STN) was performed. PD subjects were assessed under

Logistic regression four test conditions: without treatment (mof-sof), with stimulation alone (mof-son), with medication

Probabilistic neural network alone (mon-sof), and with medication and stimulation (mon-son). Principal component (PC) analysis

Support vector machine

was applied to the three components of GRF separately, where six PC scores from vertical, one from

Gait analysis

anteriorposterior and one from mediallateral were chosen by the broken stick test. Stepwise LR

analysis employed the rst two and fth vertical PC scores as input variables. Using the bootstrap

approach to compare model performances for classifying GRF patterns from normal and untreated PD

subjects, the rst three and the fth vertical PCs were attained as SVM input variables, while the same

ones plus the rst anteriorposterior were selected as PNN input variables. PNN performed better than

LR and SVM according to area under the receiver operating characteristic curve and the negative

likelihood ratio. When evaluating treatment effects, the classiers indicated that DBS-STN alone was

more effective than medication alone, but the greatest improvements occurred with both treatments

together.

& 2009 Published by Elsevier Ltd.

instead of a few parameters may improve the effectiveness of the

Parkinson disease (PD) is a neurodegenerative disorder leading analysis (Chester et al., 2007). Additionally, the correlations

to difculty in motor function, including gait and balance. Deep among variables must be considered to accurately evaluate the

brain stimulation of the subthalamic nucleus (DBS-STN) is a extent of gait abnormalities and to assess the changes resulting

treatment for advanced PD patients with disabling motor from a specic treatment (Schutte et al., 2000).

uctuations, allowing a signicant reduction in dopaminergic A clinical challenge is to understand the disease process as well

medications (Ferrarin et al., 2005). Various studies have evaluated as outcomes of potential interventions. Logistic regression (LR) is

the effects of DBS-STN using clinical motor scores (Krack et al., commonly used as a linear predictive model for diagnostic and

2003; Ostergaard and Sundae, 2006), while only a few have prognostic tasks. Recently, computational intelligence techniques

quantitatively assessed the gait of PD patients (Liu et al., 2005; such as articial neural networks (ANN) and support vector

Ferrarin et al., 2005). Gait speed is shown to be the variable most machines (SVM) have played an important role in gait classica-

affected by the DBS-STN; however, it does not take into account tion and the diagnosis of diseases (Lai et al., 2009). Studies have

atypical waveforms and therefore does not provide enough compared the predictive ability of LR and ANN (Dreiseitl and

information about the gait pattern (Schwartz and Rozumalski, Ohno-Machado, 2002; Song et al., 2005). ANN modeling has been

used in gait analysis focusing on pattern recognition (Hahn et al.,

2005), as well as for classifying normal and pathological patterns

n

Corresponding author. Tel.: +55 21 25628577; fax: + 55 21 25628591.

(Lafuente et al., 1997; Su and Wu, 2000). SVM has recently been

E-mail address: jn@peb.ufrj.br (J. Nadal). used for automated identication of gait pathologies (Begg et al.,

doi:10.1016/j.jbiomech.2009.10.018

ARTICLE IN PRESS

A.M.S. Muniz et al. / Journal of Biomechanics 43 (2010) 720726 721

2005; Lai et al., 2009). However, none of the past studies (Schumacher et al., 1996):

compared LR, probabilistic neural network (PNN) and SVM in 1

Px Pn 1

classifying gait patterns or evaluated the effect of therapeutic 1 ebo i 1

bi Pci

interventions on ground reaction force (GRF) of PD patients. where bo is the intercept and bi is the coefcient associated with the explanatory

This study evaluated LR, PNN and SVM models for discriminat- variable PCi. The maximum likelihood estimation method is used to estimate the b

ing between normal and PD subjects using principal components coefcients. For subjects classication, the PC scores were selected as the

derived from the GRF as input variables. For performance independent variables for computing the natural logarithm of the odds ratio; the

ratio between the probabilities that an event will or will not occur (1controls

evaluation, the accuracies (ACC) and the areas under the receiver

and 0PD subjects). The classication threshold was set to 0.5.

operating characteristic (ROC) curves (AUC) based on 1000

bootstrap runs of the classiers were compared. The effects of

2.4. Probabilistic neural networks

DBS-STN on GRF with and without medication were also

evaluated with both the models.

PNN is a feedforward ANN developed by Specht (1990), in which the response

to an input pattern is processed from one layer to the next, without feedback paths

to previous layers. A typical PNN has four layers: input, pattern, summation and

2. Materials and methods

output. The input units supply the same values to all pattern units. The pattern

units form a dot product between the input pattern vector x and a weight vector wi

2.1. Studied groups (zi = xwi), which is followed by the nonlinear neuron activation function:

w x0 wi x

The subjects (n= 45) consisted of 30 healthy controls (20 women) and 15 PD gzi exp i 2

2s2

patients (three women). The healthy subjects, recruiting by advertisements, were

without neurological illness, degenerative conditions or any general disease that This Bayesian function takes into account the relative likelihood of events and

might interfere with gait (Table 1). A screening questionnaire was completed to uses a priori information to improve the prediction (Specht and Romsdahl, 1994).

guarantee that subjects were suitable for the study. The PD subjects were recruited The summation units simply sum the inputs from the pattern units, corresponding

from the Parkinson Disease and Movement Disorder Center of the University of to the category from which the training patterns were selected. Repeating this

Kansas Medical Center. All PD subjects had undergone bilateral DBS-STN and were procedure for each class, the un-normalized density functions gk (x), for k= 1,

stable when the study was conducted. Each subject signed an informed consent 2, y, K were estimated. The Bayesian probability that the case was from class k is

approved by the local Institutional Review Board. as follows:

gi x

Px A k Pk 3

2.2. Experimental protocol and signal processing i1 gi x

The output units have a competitive transfer function that picks the maximum

For each PD subject, quantitative gait measurement was obtained on two of the probabilities and produces 1 for one class (normals) and 0 (PD patients) for

different days. In the rst session, the subject had taken the usual dose of PD the other.

medication and stimulators were turned on. The gait assessment was rst

conducted with both medication and stimulation (mon-son). After turning the

2.5. Support vector machine

stimulators off for 30 min, the measurements were repeated (mon-sof). In the second

session, the subjects were without medication for at least 12 h. Gait analysis was rst

conducted with stimulation (mof-son), and repeated after 30 min without stimula- The SVM estimates a function for classifying data into two classes (Vapnik,

tion (mof-sof). Due to technical problems, some subjects did not complete all tests. 2000). Using a nonlinear transformation ^ that depends on a regularization

Therefore, 13 subjects were evaluated in mof-sof, 12 in mof-son, 14 in mon-sof and 11 parameter C (Begg et al., 2005), the input vectors are placed into a high-

in mon-son conditions. Subjects from the control group were evaluated only once. dimensional feature space, where a linear separation is employed. To construct a

The quantitative analysis for the controls and PD subjects in the mof-sof condition nonlinear support vector classier, the inner product (x,y) is replaced by a kernel

was used to develop the classier models. The other three PD conditions were function K(x,y)

included in the developed models to evaluate the DBS-STN effect in PD treatment.

X

l

Two force platforms (AMTI, USA) were mounted in series at the middle of a f x sgn ai yi Kxi ; x b 4

walkway. All subjects practiced the walking trial ve times before the experiment. i1

The subjects walked barefoot at their self-selected speed and repeated the walking

where f(x) determines the membership of x. In this study, the normal subjects

trial ve times. The GRF from both force platforms were collected for 10 s at a

were labeled as 1 and PD subjects as +1. The SVM has two layers. During the

sampling frequency of 100 Hz, ltered using a low-pass Butterworth lter, with a

learning process, the rst layer selects the basis K(xi,x), i= 1,2, y, N from the given

cut-off frequency of 30 Hz, and normalized by subject body weight.

set of kernels, while the second layer constructs a linear function in the ^ space.

The averaged vertical, anteriorposterior and mediallateral components of

This is equivalent to nding the optimal hyperplane in the corresponding feature

GRF from ve walking trials were interpolated with cubic splines and re-sampled

space. The SVM algorithm can construct a variety of learning machines using

with 101 sample points according to the stance phase duration of each foot. Thus,

different kernel functions.

202 GRF samples were analyzed for each GRF component separately.

Each vertical, anteriorposterior and mediallateral GRF waveform was stored

in a matrix E with 43 rows (# of subjects) and 202 columns (# of GRF samples for 2.6. Variable selection

both right and left limbs). Principal component analysis (PCA) was applied to the

covariance matrices S (202 202) from each E, separately (Jollife, 2002).

The broken stick criterion (Jolliffe, 2002) was used for choosing the signicant

PCs of vertical, anteriorposterior and mediallateral GRF components for the

2.3. Logistic regression analysis. Moreover, to build a more accurate classier model, it was necessary to

evaluate which scores contributed to improvement in the task (Chang, 1983).

LR is a statistical modeling technique that estimates the probability of a In the LR model, a stepwise approach was used to select the input variables by

dichotomous outcome event being related to a set of explanatory variables the Akaike information criterion (AIC), followed by w2 tests to contrast with a full

model including all PC scores selected by the broken stick criterion or with subsets

of variables close to the nal model (Krzanowski, 1998).

PNN requires the selection of the optimal value for the width (s2) of the radial

Table 1 basis function. For optimizing s2 and selecting the relevant input variables, PNN

Subjects characteristics. models were trained and evaluated by the bootstrap method, considering each

possible combination of scores and varying the values s2 in the interval [0.1, 1].

Control group PD patients Briey, bootstrapping generates training sets drawing samples with replacement

from the original data set.

Age (years) 50.17 7.8 56.4 7 8.3 For the SVM, the appropriate kernel function, the number of PC scores to be

Mass (kg) 90.56 7 15.7 50.52 7 8.02 used as input variables and the parameter C were evaluated using the same

Height (m) 1.73 70.08 1.67 7 0.09 bootstrap approach applied to the PNN. The input set was determined considering

Duration of PD (years) 12.27 4.3 N.A. each possible combination of scores. All SVM models were trained over the range

Time since surgery (months) 15.17 9.5 N.A. C= {0.1, 1, 10, 100, 1000} using linear, polynomial and Gaussian kernels (Lai et al.,

2009).

ARTICLE IN PRESS

722 A.M.S. Muniz et al. / Journal of Biomechanics 43 (2010) 720726

Fig. 1. Averaged vertical ground reaction force from controls (bold line) and PD (dotted line): (a) mof-sof, (b) mof-son, (c) mon-sof and (d) mon-son conditions.

Fig. 2. Averaged anteriorposterior ground reaction force from controls (bold line) and PD (dotted line): (a) mof-sof, (b) mof-son, (c) mon-sof and (d) mon-son conditions.

2.7. Performance evaluation indices for each classier represented the average and standard error over all 1000

tests with different bootstrap samples. Comparisons among models were

performed using the one-way ANOVA, with a = 0.05 followed by the post-hoc

The models performances were assessed using the AUC and ACC. With small

Bonferroni test. The agreement between each model pair was compared using the

sample sizes, the application of a resampling technique such as the bootstrap

percent agreement and the Cohens kappa coefcients, using the algorithm

approach is recommended to estimate the performance of the classier (Sahiner et

proposed by Cardillo (2007) and interpreted according to the standard criteria

al., 2008). Both ACC and AUC were obtained using the set of samples not included

(Landis and Koch, 1977).

in the bootstrap sample that were obtained using the 0.632+ bootstrap method

(Efron and Tibshirani, 1997).

The predictive power of a classier to ascertain if a PD patient under treatment

has achieved a normal gait prole was evaluated with the negative likelihood ratio

(NLR) between false and true negatives. This ratio gives the odds that a subject 2.8. Effect of treatments

does not have symptoms of the disease, in this case indicating if the treatment has

a positive effect. For quantifying the effect of the treatments, the PC scores from PD subjects in

To perform a comparison among LR, PNN and SVM models, the same bootstrap the mon-sof, mof-son and mon-son conditions were calculated and used as inputs

samples were used for developing and testing the models. The AUC, ACC and NLR in the developed classiers.

ARTICLE IN PRESS

A.M.S. Muniz et al. / Journal of Biomechanics 43 (2010) 720726 723

Fig. 3. Averaged mediallateral ground reaction force from controls (bold line) and PD (dotted line): (a) mof-sof, (b) mof-son, (c) mon-sof and (d) mon-son conditions.

Fig. 4. (a) Area under ROC (AUC) and accuracy (ACC) of probabilistic neural network classier; and (b) average and 95% condence bands for AUC and ACC for one thousand

bootstrap training and testing samples as a function of the spread constants for the selected inputs.

3. Results normal pattern (Figs. 13), especially when both medication and

stimulation were used (Figs. 1d, 2d, 3d).

Visual inspection of three averaged GRF components from The Broken Stick test indicated that the rst six PCs from

normal and PD subjects showed the PD pattern moving toward a vertical GRF, the rst PC from anteriorposterior GRF and the rst

ARTICLE IN PRESS

724 A.M.S. Muniz et al. / Journal of Biomechanics 43 (2010) 720726

PC from mediallateral GRF, which respectively explained 91.1%, The best SVM performance was obtained for the linear kernel

69.1% and 62.1% of the total variation, should be considered in the with C = 1 and four inputs (the rst three and the fth vertical GRF

analysis. PC scores), for both AUC and ACC criteria (Table 2).

The stepwise selection of LR variables identied the rst, The resulting values of AUC, ACC and NLR for the 1000

second and fth PC scores from vertical GRF as the signicant bootstrap samples indicate that all methods presented high

inputs. Contrasting with the LR model including all eight scores performance indexes (Fig. 5). The ANOVA showed a signicant

under a w2 distribution, this model yielded a higher P value difference among these values (Table 3), indicating that PNN

(P40.05), indicating a non-signicant difference between mod- presented better AUC and NLR and the SVM a better ACC.

els, which validates the selected PC scores as input variables in the According to the standard criteria for interpreting Cohens kappa

LR classier. coefcients, the classication agreement was substantial in the

The highest PNN performance occurred with the rst three comparison between PNN and SVM, and just moderate between

and the fth vertical PCs, and the rst anteriorposterior PC LR and PNN and between LR and SVM (Table 4).

(Fig. 4a). The nal model including these ve variables was

thus maintained for comparison among model and treatment

assessment. The AUC indicated higher average performance Table 3

values within the whole range of spread constants, reaching its Average and standard error and P value for AUC, ACC and NLR.

maximum with s2 =0.8, while the ACC has the maximum with

s2 = 0.1 (Fig. 4b). To make a compromise between AUC and ACC, a LR PNN SVM P

0.1 spread was chosen. The AUC ( 7Standard Error) and the ACC at

AUC 0.9813 70.0314 0.9867 7 0.0195 0.9463 7 0.0622 o 0.001n, + ,y

this spread constant were 0.98 ( 70.019) and 0.92 ( 70.058), ACC 0.9185 70.0711 0.9273 7 0.0729 0.94607 0.0556 o 0.001n, + ,y

respectively. NLR 0.1216 70.1761 0.0354 7 0.0942 0.0569 7 0.1172 o 0.001n, + ,y

different from SVM.

Table 2

Table 4

The best SVM bootstrap training and test set area under ROC (AUC), accuracy (ACC)

Comparison between models using the Cohens kappa coefcient.

and the selected scores inputs.

Kernel C AUC ACC Variables

Linear 1 0.9463 0.9460 1, 2, 3 and 5a

LR and SVM 67 33 0.4127 Moderate agreement

Gaussian (s = 1) 10 0.8907 0.9115 1, 4 and 7a

PNN and SVM 86.5 13.5 0.8649 Substantial

Polynomial (d =3) 1 0.8979 0.9291 1, 5 and 7a

agreement

a

Variables 15 refer to the rst to fth PC scores of the vertical GRF, while LRlogistic regression, PNNprobabilistic neural network, and SVM support

variable 7 is the rst PC score of the anteriorposterior horizontal GRF. vector machine

Fig. 5. Box plot of performance indexes of LR, PNN and SVM models obtained with 1000 bootstraps: (a) AUC; (b) ACC; and (c) NLR.

ARTICLE IN PRESS

A.M.S. Muniz et al. / Journal of Biomechanics 43 (2010) 720726 725

Table 5

Comparison among logistic regression (LR), probabilistic neural network (PNN) and support vector machine (SVM) models.

Mof-son 10 (83.3%) 2 (16.7%) 4 (33.3%) 8 (66.7%) 4 (33.3%) 8 (66.7%) 12 (100%)

Mon-son 10 (90.9%) 1 (9.1%) 7 (63.3%) 4 (36.7%) 7 (63.3%) 4 (36.7%) 11 (100%)

When assessing treatments (Table 5), the classiers ranked studies that compared only the LR and ANN (Nguyen et al., 2002;

treatments in the same order, with DBS-STN presenting better Song et al., 2005).

results than just medication, and further GRF improvements with Following Landis and Koch (1977) method for describing the

combined treatments. strength of agreement associated with the Cohens kappa statistics

(Table 4), the present results indicate that PNN and LR had a

moderate agreement and PNN and SVM presented a substantial

agreement. These ndings are also in agreement with Dhana-

4. Discussion lakshmi et al. (2009), who demonstrated that SVM and ANN can

be effectively used as classication approaches. Additionally,

This is the rst study that used force platforms for assessing the these results are in line with other research (Begg et al., 2005;

effects of the DBS-STN on PD gait. Although most DBS-STN Lai et al., 2009), indicating that SVM had good performance in

assessments show an improvement in gait speed (Liu et al., 2005; pathological gait pattern recognition.

Ferrarin et al., 2005), none of them quantied its effect by analyzing When applied for PD treatment comparison, the three models

the entire waveform. GRF is one important indicator of joint moments ranked results from testing conditions in the same order

and muscle activities (Zajac et al., 2003), and is partially explained by indicating that DBS-STN alone was more effective than medication

subjects velocity or cadence (Winter, 1991). Its improvement with alone in moving PD patients to a normal GRF pattern. The best

treatment as measured by force peaks represents only a preliminary results occurred with the combination of treatments (Table 2).

step in gait analysis (Loslever and Barbier, 1998). Indeed, it does not Similarly, past studies (Ferrarin et al., 2005; Xie et al., 2001) have

allow a clear classication of subjects in different conditions. reported enhancements in gait performance with stimulation

Conversely, PCA coefcients account for information from the entire alone and further improvement when combined with medication.

GRF waveforms, rather than discrete parameters as in most past Those authors suggested a synergistic effect of DBS-STN and

studies (Li and Hamill, 2002; Hsiang and Chang, 2002), which levodopa for PD symptoms. PNN classied fewer PD subjects as

disregard the high correlation among those parameters (Chester et al., normal, presenting reduced classication difference between

2007). Additionally, PCA provided a substantial dimension reduction DBS-STN and medication alone (Table 2), agreeing with past

of classier inputs and simplied the use of the bootstrap technique, studies (Ferrarin et al., 2005; Xie et al., 2001). Moreover, PNN had

which are recommended for small sample sizes (Sahiner et al., 2008). lower NLR or odds of a subject having a normal gait pattern;

Not all PCs selected by the broken stick method are necessarily therefore, being more reliable to classify the PD subjects as

important for classication, since the best trade-off between normal when the GRF pattern is improved.

variance and classication should be taken into account (Chang,

1983; Jolliffe, 2002). Dreiseitl and Ohno-Machado (2002) also

recommend determining which variables are relevant to reach a 5. Conclusion

good model performance. The stepwise and bootstrap approaches

allowed for the selection of the PCs that maximized performance The LR, PNN and SVM models presented high performance

of the three tested models, using an adequate criteria. The indexes for classifying GRF pattern of normal subjects and

comparison using AUC and ACC also allowed for the selection of untreated PD. When using the bootstrap approach, PNN per-

the best spread value in PNN, as recommended by Hernandez- formed better according to AUC and NLR criteria and the SVM

Caraballo et al. (2005), as well as the best kernel and C parameter showed the best ACC.

in SVM. When evaluating the effect of treatments, the three classiers

The bootstrap is recognized as a robust approach for estimat- indicated DBS-STN alone was more effective than medication

ing classier performance under a limited sample size (Sahiner et alone, with the greatest improvement with combined treatments.

al., 2008). According to Sahiner et al. (2008), the 0.632+ bootstrap However, PNN was more restrictive for accepting the patients GRF

approach provides the lowest bias with the greatest accuracy. In as normal.

the present study, with 1000 bootstrap samples, the PNN

presented the highest AUC and the lowest NLR values compared

to LR and SVM, while SVM had the highest ACC value (Table 3). Conict of interest statement

AUC has been recommended as better performance index, since it

presents higher convergence than ACC (Brandley, 1997), and None of the authors have any nancial or personal relationship

represents the average sensitivity across all possible specicities that could inappropriately inuence the work submitted for

(Glas et al., 2003). On the other hand, ACC is still the most publication.

commonly used index, although dependent on the prevalence of

the disorder in the studied group (Glas et al., 2003). The NLR

describes the discriminatory properties of negative test results. Acknowledgements

According to Deeks (2009), the NLR values correspond to strong

diagnostic evidence. Thus, PNN was more rigorous than LR and This work was partially supported by the Brazilian Research

SVM in accepting a pattern as normal, which agreed with other Council (CNPq) and by CAPES and FAPERJ foundation.

ARTICLE IN PRESS

726 A.M.S. Muniz et al. / Journal of Biomechanics 43 (2010) 720726

References Lai, D.T., Levinger, P., Begg, R.K., Gilleard, W.L., Palaniswami, M., 2009. Automatic

recognition of gait patterns exhibiting patellofemoral pain syndrome using a

support vector machine approach. IEEE Transactions on Information Technol-

Brandley, A.P., 1997. The use of the area under the roc curve in the evaluation of ogy in Biomedicine 13 (4), 810817.

machine learning algorithms. Pattern Recognition 30 (7), 11451159. Landis, J.R., Koch, G.G., 1977. The measurement of observer agreement for

Begg, R.K., Palaniswami, M., Owen, B., 2005. Support vector machines for categorical data. Biometrics 33 (1), 159174.

automated gait classication. IEEE Transactions on Biomedica Engineering 53 Li, L, Hamill, J., 2002. Characteristics of the vertical ground reaction force

(5), 828838. component prior to gait transition. Research Quarterly for Exercise and Sport

Cardillo G., 2007. Cohens kappa: compute the Cohens kappa ratio on a 2x2 matrix. 73 (3), 229237.

http://www.mathworks.com/matlabcentral/leexchange/15365. Liu, W., Mcintire, K., Kim, S.H., Zhang, J., Dascalos, S., Lyons, K.E., Pahwa, R., 2005.

Chang, W.C., 1983. On using principal components before separating a mixture of Quantitative assessments of the effect of bilateral subthalamic stimulation on

two multivariate normal distribuitions. Applied Statistics 32 (3), 267275. multiple aspects of sensoriomotor function for patients with Parkinsons

Chester, W.L., Tingley, M., Biden, E.N., 2007. An extended index to quantify disease. Parkinsonism and Related Disordens 11, 503508.

normality of gait in children. Gait and Posture 25, 549554. Loslever, P., Barbier, F., 1998. Multivariate graphical presentation for gait

Dhanalakshmi, P., Palanivel, S., Ramalingam, V., 2009. Classication of audio rehabilitation study. Gait and Posture 7, 3944.

signals using SVM and RBFNN. Expert Systems with Applications 36, 6069 Nguyen, T., Malley, R., Inkelis, S., Kuppermann, N., 2002. Comparison of prediction

6075. models for adverse outcome in pediatric meningococcal disease using articial

Deeks, J.J., 2009. Systematic reviews in health care Systematic reviews of neural netweork and logistic regression analyses. Journal of Clinical Epide-

evaluations of diagnostic and screening tests. British Medical Journal 323, miology 55, 687695.

157162. Ostergaard, K., Sundae, N.A., 2006. Evolution of Parkinsons disease during 4 years

Dreiseitl, S., Ohno-Machado, L., 2002. Logistic regression and articial neural of bilateral deep brain stimulation of the subthalamic nucleus. Movement

network classication models: a methodology review. Journal of Biomedical Disorders 21 (5), 624631.

Informatics 35, 352359. Sahiner, B., Chan, H., Hadjiiski, L., 2008. Classier performance prection for

Efron, B., Tibshirani, R., 1997. Improvements on cross-validation: The .632 + computer-aided diagnosis using a limited dataset. Medical Physics 35 (4),

bootstrap method. Journal of the American Statistical Association 92 (438), 15591570.

548560. Schumacher, M., Robner, R., Wemer, V., 1996. Neural networks and logistic

Ferrarin, M., Rizzone, M., Bergamasco, B., Lanotte, M., Recalcati, M., Pedotti, A., regression. Part I Computational Statistics and Data Analysis 21, 661682.

Lopiano, L., 2005. Effects of bilateral subthalamic stimulation on gait Schutte, L.M., Narayanan, U., Stout, J.L., Selber, P., Gage, J.R., 2000. An index for

kinematics and kinetics in Parkinsons disease. Experimental Brain Research quantifying deviations from normal gait. Gait and Posture 11, 2531.

160, 517527. Schwartz, M.H., Rozumalski, A., 2008. The gait deviation index: A new

Glas, A.S., Lijmer, J.G., Prins, M.H., Bonsel, G.J., Bossuyt, P.M.M., 2003. The diagnostic comprehensive index of gait pathology. Gait and Posture 28, 351357.

odds ratio: a single indicator of test performance. Journal of Clinical Song, J.H., Venkatesh, S.S., Conant, E.A., Arger, P.H., Sehgal, Chandra M., 2005.

Epidemiology 56, 11291135. Comparative analysis of logistic regression and articial neural network

Hahn, M.E., Farley, A.M., Lin, V., Chou, L., 2005. Neural network estimation of for computer-aided diagnosis of breast masses. Academic Radiology 12,

balance control during locomotion. Journal of Biomechanics 38, 717724. 487495.

Hernandez-Caraballo, E.A., Rivas, F., Perez, A.G., Marco-Parra, M., 2005. Evaluation Specht, D.F., 1990. Probabilistic neural networks. Neural Networks 3, 109118.

of chemometric techniques and articial neural networks for cancer screening Specht, D., Romsdahl, H., 1994. Experience with adaptive probabilistic neural

using Cu, Fe, Se and Zn concentrations in blood serum. Analytica Chimica Acta networks and adaptive general regression neural networks. Neural Networks

533, 161168. 12031208.

Hsiang, S, Chang, C., 2002. The effect of gait speed and load carrying on the Su, F., Wu, W., 2000. Design and testing of a genetic algorithm neural network in

reliability of ground reaction forces. Safety Science 40, 639657. the assessment of gait patterns. Medical Engineering and Physics 22, 6774.

Jolliffe, I.T., 2002. In: Principal Component Analysis second ed. Springer, New York. Vapnik, VN., 2000. In: The Nature of Statistical Learning Theory second ed.

Krack, P., Batir, A., Van Blercom, N., Chabardes, S., Fraix, V., Ardouin, A., et al., 2003. Springer, New York.

Five-year follow-up of bilateral stimulation of the subthalamic nucleus in Winter, D.A., 1991. In: The Biomechanics and Motor Control of Human Gait:

advanced parkinsons disease. The New England Journal of Medicine 340 (20), Normal, Elderly and Pathological second ed. University of Waterloo Press,

19251934. Waterloo.

Krzanowski, W.J., 1998. In: An Introduction to Statistical Modelling. Arnold, Xie, J., Krack, P., Benabid, A., Pollak, P., 2001. Effect of bilateral subthalamic nucleus

London. stimulation on parkinsonian gait. Journal of Neurology 248, 10681072.

Lafuente, R., Belda, J.M., Sanchez-Lacuesta, J., Soler, C., Prat, J., 1997. Design and test Zajac, F.E., Sachin, R.R., Kautz, S.A., 2003. Biomechanics and muscle coordination of

of neural networks and statistical classiers in computer-aided movement human walking part II: lessons from dynamical simulations and clinical

analysis: a case study on gait analysis. Clinical Biomechanics 13 (3), 216229. implications. Gait and Posture 17, 117.

- A Real-Time Interference Monitoring Technique for GNSS Based on a Twin Support Vector Machine Method.pdfUploaded byfabry8080
- IJEST Special Issue CompleteUploaded byRudra Kumar Mishra
- Mcq MockUploaded byRohit Sohlot
- R InstallationUploaded bymanishscry
- Iaetsd-jaras-load Reduction in Videos Using Advanced FastUploaded byiaetsdiaetsd
- Art. Redes Neuronales de BiomasaUploaded byRaquel Tenorio
- Fully Automatic Facial Action Unit Detection and Temporal Analysis.pdfUploaded bySusan Thomas
- Articulated Pose Estimation With Flexible Mixtures-Of-partsUploaded byJosé Alejandro Mesejo Chiong
- A Review on Construction Cost Forecasting TechniquesUploaded byIRJET Journal
- 2532511Uploaded bynaghito
- svmUploaded byYogender Kumar Gupta
- Fath on y 2017 AdversarialUploaded byMochamad Asryl Aziz
- IRJET-Intelligent Emotion Detection System using Facial ImagesUploaded byIRJET Journal
- Hand gesture recognition using support vector machineUploaded bytheijes
- AN EFFICIENT HEMORRHAGE DETECTION USING SVM CLASSIFIERUploaded bySeun -nuga Daniel
- documentUploaded bySheik Mohamed
- Super Parsing Sentiment Classification With Review ExtractionUploaded byAboMallik
- research proposal machine learningUploaded byazmat_alikhan
- 2D 3D Face RecognitionUploaded bySatya Murty
- en_Tanagra_KMO_Bartlett.pdfUploaded bythe
- pdf to word.docxUploaded byradhika
- Winfree Etal 2009 EcologyUploaded byJosé Crespo
- Genstat 5.32Uploaded byMahmud ornob
- A SHORT ACCULTURATION SCALE FOR MEXICAN-AMERICAN POPULATIONS.pdfUploaded byA. A. Suleiman
- paper_89Uploaded bysaudade96
- Modeling and advanced control method of PVC polymerizationUploaded bySyed Ubaid Ali
- BHS_FinalArticleForConferenceUploaded byFrances Neele
- 1853-automatic-choice-of-dimensionality-for-pca.pdfUploaded byBilly
- METHODOLOGICAL STUDY OF OPINION MINING AND SENTIMENT ANALYSIS TECHNIQUESUploaded byijsc
- IJCTT-V3I2P111Uploaded bysurendiran123

- ahroomUploaded byseyedmohamad
- yes.pdfUploaded bygren900
- Ferber 2016Uploaded byseyedmohamad
- Mrj v8n1p72 FaUploaded byseyedmohamad
- 3000813920314Uploaded byseyedmohamad
- 967-5273-1-SPUploaded byseyedmohamad
- Par Des 2016Uploaded byseyedmohamad
- 114-F.pdfUploaded byseyedmohamad
- Alcock 2016Uploaded byseyedmohamad
- Terms of CollaborationUploaded byseyedmohamad
- Pi is 0021929015006387Uploaded byseyedmohamad
- Everything You Need to Know About SquattingUploaded byנתנאל מושקוביץ
- Pi is 0021929014006836Uploaded byseyedmohamad
- Pi is 0021929002000465Uploaded byseyedmohamad
- Pi is 0021929015004169Uploaded byseyedmohamad
- Pi is 002192901400013 xUploaded byseyedmohamad
- MiladUploaded byseyedmohamad
- Pi is 0021929005004392Uploaded byseyedmohamad
- 114-FUploaded byseyedmohamad
- Pi is 0021929015005291Uploaded byseyedmohamad

- Groupe ThreeUploaded byHAFTU TEKLAY
- Lecture 16Uploaded bysureshrnal
- htkbook2Uploaded byBB
- Mhasan,Predicting 28 Days Compressive Strength of Concrete From 7 Days Test ResultUploaded byNeven Ahmed Hassan
- Math117FinalReview-Fall2014Uploaded bySayeesh Kapu
- Mcq Introduction With Correct AnswersUploaded byKranti Prajapati
- EmotionalandsexualwellnessandqualityoflifeinwomenwithRokitanskysyndromeUploaded bynhsajib
- Decision Under Uncertainty Lesson 2 NewUploaded byDivya Makwana
- normality skewness kurtosisUploaded byapi-285777244
- Borrego-Douglas-Amelink-Quantitative-Qualitative-and-Mixed-Research-Methods-in-Engineering-Education.pdfUploaded byanon_335503439
- Bio Stat 601Uploaded bySocks
- 1_tests of HypothesesUploaded bychristine
- discrete random variablesUploaded bykentigernk
- Piazzesi Swanson(JME2008)Uploaded bysbastonvilla
- Sample Research Paper for Ch1, ch3, ch4 and ch 5.docxUploaded byMila Domingo
- Job Satisfaction and PromotionsUploaded byIrum Mehvish
- hea-31-1-87Uploaded byAiiyuu Lesthari Dhian
- ProceedUploaded bykapil1248
- How to use SPSSUploaded bydhimba
- C. R. RaoUploaded bydroy21
- 11 -July, 2011-R4.pdfUploaded bySurendra Yadav
- Math MistakesUploaded byValda Paužuolytė
- StrengthPerviousUploaded byEtana O-o
- Chapter 3Uploaded byhungthuhienbk
- poster_programmeUploaded byDarwin Gomez
- Bootstrap for Panel Data (Ppt), HounkannounonUploaded byanon_183065072
- FAQs Unit 01 and 02 MBF103Uploaded byAjeet Verma
- Financial Ratio Selection for Business Failure Prediction Using Soft Set TheoryUploaded byfirebirdshockwave
- vol1.1Jyoti_PohaneUploaded byVienna Corrine Q. Abucejo
- 498-1476-1-PBUploaded byCharleen Joyce Usacdin