A Computational Framework for Interpretable Anomaly Detection and Classification of Multivariate Time Series with Application to Human Gait Data Analysis

Sensor-based methods for human gait analysis often utilize electromyography capturing rich time-series data. Then, for transparent and explainable analysis interpretable methods are of prime importance. This paper presents analytical approaches in a framework for interpretable anomaly detection and classification of multivariate time series for human gait analysis. We exemplify the application utilizing a real-world medical dataset in the biomechanical orthopedics domain.

[1]  Martin Atzmüller,et al.  Declarative Aspects in Explicative Data Mining for Computational Sensemaking , 2017, DECLARE.

[2]  Frank Puppe,et al.  Application and Evaluation of a Medical Knowledge System in Sonography (SONOCONSULT) , 2008, ECAI.

[3]  Eamonn J. Keogh,et al.  HOT SAX: efficiently finding the most unusual time series subsequence , 2005, Fifth IEEE International Conference on Data Mining (ICDM'05).

[4]  E. Shortliffe,et al.  Clinical Decision Support in the Era of Artificial Intelligence. , 2018, JAMA.

[5]  M. Wimmer,et al.  Kinematic evaluation of cruciate-retaining total knee replacement patients during level walking: a comparison with the displacement-controlled ISO standard. , 2009, Journal of biomechanics.

[6]  Randolph A. Miller,et al.  Review: Medical Diagnostic Decision Support Systems - Past, Present, And Future: A Threaded Bibliography and Brief Commentary , 1994, J. Am. Medical Informatics Assoc..

[7]  Martin Atzmüller,et al.  The Mining and Analysis Continuum of Explaining Uncovered , 2010, SGAI Conf..

[8]  Heba M. Lakany,et al.  Extracting a diagnostic gait signature , 2008, Pattern Recognit..

[9]  Frank Puppe,et al.  Semi-automatic learning of simple diagnostic scores utilizing complexity measures , 2006, Artif. Intell. Medicine.

[10]  Guy Lapalme,et al.  A systematic analysis of performance measures for classification tasks , 2009, Inf. Process. Manag..

[11]  Benjamin Klöpper,et al.  Explanation-aware feature selection using symbolic time series abstraction: Approaches and experiences in a petro-chemical production context , 2017, 2017 IEEE 15th International Conference on Industrial Informatics (INDIN).

[12]  E. Shortliffe.  Computer programs to support clinical decision making. , 1990, JAMA.

[13]  Georgiana Ifrim,et al.  Interpretable time series classification using linear models and multi-resolution multi-domain symbolic representations , 2019, Data Mining and Knowledge Discovery.

[14]  Geeta Sikka,et al.  Recent Techniques of Clustering of Time Series Data: A Survey , 2012 .

[15]  Tim Oates,et al.  Time series anomaly discovery with grammar-based compression , 2015, EDBT.

[16]  Andreas W. Kempa-Liehr,et al.  Time Series FeatuRe Extraction on basis of Scalable Hypothesis tests (tsfresh - A Python package) , 2018, Neurocomputing.

[17]  Vipin Kumar,et al.  Comparative Evaluation of Anomaly Detection Techniques for Sequence Data , 2008, 2008 Eighth IEEE International Conference on Data Mining.

[18]  Nicolas Vuillerme,et al.  EMG normalization method based on grade 3 of manual muscle testing: Within- and between-day reliability of normalization tasks and application to gait analysis. , 2018, Gait & posture.

[19]  Keleş Selami,et al.  Classification Of Emg Signals Using Decision Tree Methods , 2012 .

[20]  Frank Puppe,et al.  Subgroup Mining for Interactive Knowledge Refinement , 2005, AIME.

[21]  Andreas W. Kempa-Liehr,et al.  Distributed and parallel time series feature extraction for industrial big data applications , 2016, ArXiv.

[22]  T. Warren Liao,et al.  Clustering of time series data - a survey , 2005, Pattern Recognit..

[23]  Frank Puppe,et al.  Systematic introduction to expert systems - knowledge representations and problem-solving methods , 2011 .

[24]  Ying Wah Teh,et al.  Time-series clustering - A decade review , 2015, Inf. Syst..

[25]  F. Puppe,et al.  Profiling Examiners using Intelligent Subgroup Mining , 2005 .

[26]  M. Wimmer,et al.  TKA patients with unsatisfying knee function show changes in neuromotor synergy pattern but not joint biomechanics. , 2017, Journal of electromyography and kinesiology : official journal of the International Society of Electrophysiological Kinesiology.

[27]  Eamonn J. Keogh,et al.  The great time series classification bake off: a review and experimental evaluation of recent algorithmic advances , 2016, Data Mining and Knowledge Discovery.

[28]  G. J. Nalepa,et al.  From Context Mediation to Declarative Values and Explainability , 2018 .

[30]  Eamonn J. Keogh,et al.  On the Need for Time Series Data Mining Benchmarks: A Survey and Empirical Demonstration , 2002, Data Mining and Knowledge Discovery.

[31]  Martin Atzmüller,et al.  Feature-Set-Engineering for Detecting Freezing of Gait in Parkinson's Disease using Deep Recurrent Neural Networks , 2019, ArXiv.