A variational method to determine the most representative shape of a set of curves and its application to knee kinematic data for pathology classification

This study investigates a variational method to determine the most representative shape of a set of knee kinematic curves with application to knee pathology classification. Although they provide essential information for pathology classification, knee kinematic curves are characterized by high intra-class variability and outliers are often present. As a result, a set of several measurement curves are acquired of any single individual which are then averaged before their use for pathology classification. Rather than using the average of an individual’s recorded measurement curves, this method determines a better representative curve by first correcting the data to account for outliers occurrence and class variability using a variational method. The correction is performed by simultaneous minimization of a set of objective functions, one for each curve in the measurement set, and consisting of a weighed sum of two terms: a data term of conformity of the corrected curve to the given curve, and a regularization term of proximity of the corrected curve to the mean of all the corrected curves to inhibit the influence of outliers in the set. Validation tests were performed to discriminate between knee osteoarthritis data (OA) and non-OA data. Using a support vector machine, the classification accuracy with the proposed representation was 86%, with 81% sensitivity and 90% specificity, compared to 83% accuracy for the standard representation by average, with 76% sensitivity and 90% specificity. The representation has also been tested within the OA category to distinguish the femero-tibial patholgy from the femero-patellar, giving 76% accuracy, with 76% sensitivity and 76% specificity, compared to 69% accuracy, with 62% sensitivity and 76% specificity. These significant improvements by the proposed method warrant its further investigation by application to other biomedical engineering pattern classification problems and datasets.

Download Full-text

A Sample-Encoding Generalization of the Kohonen Associative Memory and Application to Knee Kinematic Data Representation and Pathology Classification

Applied Sciences ◽

10.3390/app9091741 ◽

2019 ◽

Vol 9 (9) ◽

pp. 1741 ◽

Cited By ~ 3

Author(s):

Badreddine Ben Nouma ◽

Amar Mitiche ◽

Neila Mezghani

Keyword(s):

Three Dimensional ◽

Data Representation ◽

Small Sample ◽

Repeated Measurements ◽

Training Data ◽

Multidimensional Data ◽

Knee Kinematic ◽

Kinematic Data ◽

Pathology Classification ◽

Knee Pathology

Knee kinematic data consist of a small sample of high-dimensional vectors recording repeated measurements of the temporal variation of each of the three fundamental angles of knee three-dimensional rotation during a walking cycle. In applications such as knee pathology classification, the notorious problems of high-dimensionality (the curse of dimensionality), high intra-class variability, and inter-class similarity make this data generally difficult to interpret. In the face of these difficulties, the purpose of this study is to investigate knee kinematic data classification by a Kohonen neural network generalized to encode samples of multidimensional data vectors rather than single such vectors as in the standard network. The network training algorithm and its ensuing classification function both use the Hotelling T 2 statistic to evaluate the underlying sample similarity, thus affording efficient use of training data for network development and robust classification of observed data. Applied to knee osteoarthritis pathology discrimination, namely the femoro-rotulian (FR) and femoro-tibial (FT) categories, the scheme improves on the state-of-the-art methods.

Download Full-text

3D Kinematics and Decision Trees to Predict the Impact of a Physical Exercise Program on Knee Osteoarthritis Patients

Applied Sciences ◽

10.3390/app11020834 ◽

2021 ◽

Vol 11 (2) ◽

pp. 834

Author(s):

Marwa Mezghani ◽

Nicola Hagemeister ◽

Youssef Ouakrim ◽

Alix Cagnin ◽

Alexandre Fuentes ◽

...

Keyword(s):

Physical Exercise ◽

Knee Osteoarthritis ◽

Decision Trees ◽

Exercise Program ◽

Patient Specific ◽

Pain Level ◽

Knee Kinematic ◽

Kinematic Data ◽

3D Kinematics ◽

The Impact

Measuring knee biomechanics provides valuable clinical information for defining patient-specific treatment options, including patient-oriented physical exercise programs. It can be done by a knee kinesiography test measuring the three-dimensional rotation angles (3D kinematics) during walking, thus providing objective knowledge about knee function in dynamic and weight-bearing conditions. The purpose of this study was to assess whether 3D kinematics can be efficiently used to predict the impact of a physical exercise program on the condition of knee osteoarthritis (OA) patients. The prediction was based on 3D knee kinematic data, namely flexion/extension, adduction/abduction and external/internal rotation angles collected during a treadmill walking session at baseline. These measurements are quantifiable information suitable to develop automatic and objective methods for personalized computer-aided treatment systems. The dataset included 221 patients who followed a personalized therapeutic physical exercise program for 6 months and were then assigned to one of two classes, Improved condition (I) and not-Improved condition (nI). A 10% improvement in pain was needed at the 6-month follow-up compared to baseline to be in the improved group. The developed model was able to predict I and nI with 84.4% accuracy for men and 75.5% for women using a decision tree classifier trained with 3D knee kinematic data taken at baseline and a 10-fold validation procedure. The models showed that men with an impaired control of their varus thrust and a higher pain level at baseline, and women with a greater amplitude of internal tibia rotation were more likely to report improvements in their pain level after 6 months of exercises. Results support the effectiveness of decision trees and the relevance of 3D kinematic data to objectively predict knee OA patients’ response to a treatment consisting of a physical exercise program.

Download Full-text

Pattern Classification by the Hotelling Statistic and Application to Knee Osteoarthritis Kinematic Signals

Machine Learning and Knowledge Extraction ◽

10.3390/make1030045 ◽

2019 ◽

Vol 1 (3) ◽

pp. 768-784

Author(s):

Badreddine Ben Nouma ◽

Amar Mitiche ◽

Youssef Ouakrim ◽

Neila Mezghani

Keyword(s):

Knee Osteoarthritis ◽

Reference Data ◽

Small Sample ◽

Repeated Measurements ◽

Kinematic Data ◽

High Data ◽

Current State ◽

Pathology Classification ◽

Discrete Curves ◽

Knee Pathology

The analysis of knee kinematic data, which come in the form of a small sample of discrete curves that describe repeated measurements of the temporal variation of each of the knee three fundamental angles of rotation during a subject walking cycle, can inform knee pathology classification because, in general, different pathologies have different kinematic data patterns. However, high data dimensionality and the scarcity of reference data, which characterize this type of application, challenge classification and make it prone to error, a problem Duda and Hart refer to as the curse of dimensionality. The purpose of this study is to investigate a sample-based classifier which evaluates data proximity by the two-sample Hotelling T 2 statistic. This classifier uses the whole sample of an individual’s measurements for a better support to classification, and the Hotelling T 2 hypothesis testing made applicable by dimensionality reduction. This method was able to discriminate between femero-rotulian (FR) and femero-tibial (FT) knee osteoarthritis kinematic data with an accuracy of 88.1 % , outperforming significantly current state-of-the-art methods which addressed similar problems. Extended to the much harder three-class problem involving pathology categories FR and FT, as well as category FR-FT which represents the incidence of both diseases FR and FT in a same individual, the scheme was able to reach a performance that justifies its further use and investigation in this and other similar applications.

Download Full-text

A Combined Visualization Method for Multivariate Data Analysis. Application to Knee Kinematic and Clinical Parameters Relationships

Applied Sciences ◽

10.3390/app10051762 ◽

2020 ◽

Vol 10 (5) ◽

pp. 1762

Author(s):

Fatima Bensalma ◽

Glen Richardson ◽

Youssef Ouakrim ◽

Alexandre Fuentes ◽

Michael Dunbar ◽

...

Keyword(s):

Canonical Correlation ◽

Multivariate Data Analysis ◽

Multivariate Data ◽

Demographic Characteristics ◽

Clinical Parameters ◽

Visualization Method ◽

Knee Kinematic ◽

Kinematic Data ◽

Analysis Application ◽

Multivariate Relationship

This paper aims to analyze the correlation structure between the kinematic and clinical parameters of an end-staged knee osteoarthritis population. The kinematic data are a set of characteristics derived from 3D knee kinematic patterns. The clinical parameters include the answers of a clinical questionnaire and the patient’s demographic characteristics. The proposed method performs, first, a regularized canonical correlation analysis (RCCA) to evaluate the multivariate relationship between the clinical and kinematic datasets, and second, a combined visualization method to better understand the relationships between these multivariate data. Results show the efficiency of using different and complementary visual representation tools to highlight hidden relationships and find insights in data.

Download Full-text