An automated hierarchical gait pattern identification tool employing cross-correlation-based feature extraction and recurrent neural network based classification

Expert Systems (Impact Factor: 0.76). 05/2009; 26(2):202-217. DOI: 10.1111/j.1468-0394.2009.00479.x
Source: DBLP


In this paper Elman's recurrent neural network (ERNN) is employed for automatic identification of healthy and pathological gait and subsequent diagnosis of the neurological disorder in pathological gaits from the respective gait patterns. Stance, swing and double support intervals (expressed as percentages of stride) of 63 subjects were analysed for a period of approximately 300 s. The relevant gait features are extracted from cross-correlograms of these signals with corresponding signals of a reference subject. These gait features are used to train modular ERNNs performing binary and tertiary classifications. The average accuracy of binary classifiers is obtained as 90.6%–97.8% and that of tertiary classifiers is 89.8%. Hence, two hierarchical schemes are developed each of which uses more than one modular ERNN to segregate healthy, Parkinson's disease, Huntington's disease and amyotrophic lateral sclerosis subjects. The average testing performances of the schemes are 83.8% and 87.1%.

1 Follower
19 Reads
  • Source
    • "cross-correlation technique has been conveniently used in many applications like biomedical signal processing [19] [20] [21], image processing [22], robotics and remotesensing , sonar and radar systems and in several other domains [23] [24]. In [19] [20] cross-correlation technique has been successfully used for pattern recognition of gait and EEG signals respectively. One of the important achievements of the present work is the successful use of cross-correlation technique in frequency domain for the analysis of ECG beats. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This work describes the development of a computerized medical diagnostic tool for heart beat categorization. The main objective is to achieve an accurate, timely detection of cardiac arrhythmia for providing appropriate medical attention to a patient. The proposed scheme employs a feature extractor coupled with an Artificial Neural Network (ANN) classifier. The feature extractor is based on cross-correlation approach, utilizing the cross-spectral density information in frequency domain. The ANN classifier uses a Learning Vector Quantization (LVQ) scheme which classifies the ECG beats into three categories: normal beats, Premature Ventricular Contraction (PVC) beats and other beats. To demonstrate the generalization capability of the scheme, this classifier is developed utilizing a small training dataset and then tested with a large testing dataset. Our proposed scheme was employed for 40 benchmark ECG files of the MIT/BIH database. The system could produce classification accuracy as high as 95.24% and could outperform several competing algorithms.
    Full-text · Article · Dec 2011 · Measurement
  • [Show abstract] [Hide abstract]
    ABSTRACT: The automatic classification of English fricatives from continuous speech using fuzzy logic is discussed. Two algorithms that classify the fricatives into four and six classes, respectively, were developed. This classification is achieved by combining distinctive features that are described by fuzzy sets and their associated membership functions. The two algorithms are independent; one is used for speech bandlimited to 3.2 kHz and the other for speech bandlimited to 7.5 kHz. The results obtained with the algorithms were very promising, an overall recognition rate of 89.7% being achieved for the narrowband algorithm and 76.2% for the wideband algorithm. This compares favourably to the results of perception tests carried out under the same conditions, where an overall recognition rate of 74.5% was achieved with narrowband speech and 77.6% for wideband speech. The results indicate that contextual information plays an extremely important part in the recognition of fricatives
    No preview · Conference Paper · Jul 1989
  • [Show abstract] [Hide abstract]
    ABSTRACT: A mechanically stable association between the membrane bilayer and the underlying membrane associated skeleton is important for maintaining the integrity of the plasma membrane. This is particularly true for red blood cells, which must maintain their physical integrity and deformability within the dynamic environment of the vasculature. To probe the molecular mechanisms that account for membrane stability, we have combined a magnetic force transducer with fluorescence imaged microdeformation to measure the forces required to separate the membrane bilayer by pulling thin lipid cylinders, or tethers, off of the cell surface and observed the response of integral membrane proteins to bilayer separation. The results from these experiments indicate that, over a period of up to 30 minutes, the tethering force decreases until it reaches an equilibrium value of 54.7±8.3 pN while the surface density of fluorescently labeled integral proteins on the cell body is increased after tether formation
    No preview · Conference Paper · Feb 1999
Show more