Using transfer learning for classification of gait pathologies

Abstract

Different diseases can affect an individual's gait in different ways and, therefore, gait analysis can provide important insights into an individual's health and well-being. Currently, most systems that perform gait analysis using 2D video are limited to simple binary classification of gait as being either normal or impaired. While some systems do perform gait classification across different pathologies, the reported results still have a considerable margin for improvement. This paper presents a novel system that performs classification of gait across different pathologies, with considerably improved results. The system computes the walking individual's silhouettes, which are computed from a 2D video sequence, and combines them into a representation known as the gait energy image (GEI), which provides robustness against silhouette segmentation errors. In this work, instead of using a set of handcrafted gait features, feature extraction is done using the VGG-19 convolutional neural network. The network is fine-tuned to automatically extract the features that best represent gait pathologies, using transfer learning. The use of transfer learning improves the classification accuracy while avoiding the need of a very large training set, as the network is pre-trained for generic image description, which also contributes to a better generalization when tested across different datasets. The proposed system performs the final classification using linear discriminant analysis (LDA). Obtained results show that the proposed system outperforms the state-of-the-art, achieving a classification accuracy of 95% on a dataset containing gait sequences affected by diplegia, hemiplegia, neuropathy and Parkinson's disease, along with normal gait sequences.

Computer-Vision Based Diagnosis of Parkinson’s Disease via Gait: A Survey

“…In order to guarantee the usefulness of class discrimination, Linear/Fischer discriminant analysis (LDA/ FDA) [50], [78], [94] is used that maximizes the data classification (increase intercluster distances and reduce intracluster distances) based on the idea of separating two classes by finding the linear combination of variables. Another method such as KPCA [76], an extension of PCA is based on a kernel that performs non-linear mapping to reduce the dimensionality of the feature set and decreases the computational complexity to a greater extent.…”

“…Therefore, an improved paradigm named as deep learning that runs through manifold abstraction levels in data has been designed to make updations in ANN. Deep learning techniques [94], [97] are the set of machine learning algorithms to deal potentially with the growing amount of data as more the volume of data, higher is the accuracy. The greater depth of the network automatically enhances the performance of the system.…”

“…The use of a large number of features for evaluation purpose unnecessarily increases the feature space and often affects the recognition capability of the system. In spite, existing studies have used various dimensionality reduction algorithms including PCA [93], [104], LDA [78], [94], etc. for selection of optimized features for PD diagnosis but still need efforts to improve the system's efficiency.…”

A Transfer Learning Architecture Based on a Support Vector Machine for Histopathology Image Classification

“…Different transfer learning techniques are classified in [6] into feature representation transfer, including cross-domain knowledge transfer and cross-view knowledge transfer, and classifier-based knowledge transfer, including SVM-based, TrAdaboost, and generative models. The employment of deep transfer learning is prevalent due to its superior performance and flexibility [7][8][9][10][11].…”

A Spatiotemporal Deep Learning Approach for Automatic Pathological Gait Classification

“…They use Convolutional Neural Networks (CNNs), as these architectures are able to learn visual features with a high level of abstraction. For instance, in [ 27 ], pathological gait was analyzed using the VGG-19 architecture [ 28 ], pretrained with a subset of the ImageNet database [ 29 ]. Transfer learning was used to repurpose the model for gait classification using the GEIs computed from the INIT gait dataset sequences [ 15 ].…”