Fault detection of continuous glucose measurements based on modified k-medoids clustering algorithm

As continuous glucose monitoring (CGM) systems provide critical feedback information of blood glucose concentration to the artificial pancreas for patients with type 1 diabetes (T1D), faults in CGM may seriously affect the computation of insulin infusion rates which can lead to fatal consequences accompany with hypoglycemia or hyperglycemia. In the present work, the k-medoids clustering algorithm is modified by calculating cluster number with a Bayesian Information Criterion (BIC)-based cost function and the SAC (SSE-ASW Criterion) evaluation coefficient which considers both SSE (Sum of Square due to Error) and ASW (Average Silhouette Width) criteria. Then, the modified k-medoids clustering algorithm is proposed to detect sensor failures online with CGM measurements. Different from the qualitative model-based methods and quantitative model-based methods, sufficient clean data are the only requirement of the proposed method. During online monitoring, the new glycemic variability is then tracked against predefined confidence limits during training period to indicate abnormality. The feasibility of the proposed method is successfully assessed using CGM data collected from the UVa/Padova metabolic simulator.

[1]  Youxian Sun,et al.  A classification-based fault detection method for Continuous glucose monitoring (CGM) , 2016, 2016 12th World Congress on Intelligent Control and Automation (WCICA).

[2]  Hong Zhao,et al.  An Automatic Denoising Method with Estimation of Noise Level and Detection of Noise Variability in Continuous Glucose Monitoring , 2016 .

[3]  Hong Zhao,et al.  An automatic glucose monitoring signal denoising method with noise level estimation and responsive filter updating , 2018, Biomed. Signal Process. Control..

[4]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part II: Qualitative models and search strategies , 2003, Comput. Chem. Eng..

[5]  Øyvind Stavdahl,et al.  Fault detection in glucose control: Is it time to move beyond CGM data? , 2018 .

[6]  Ali Cinar,et al.  Artificial Pancreas Systems: An Introduction to the Special Issue , 2018 .

[7]  Adrian E. Raftery,et al.  How Many Clusters? Which Clustering Method? Answers Via Model-Based Cluster Analysis , 1998, Comput. J..

[8]  A. Facchinetti,et al.  Modeling Transient Disconnections and Compression Artifacts of Continuous Glucose Sensors. , 2016, Diabetes technology & therapeutics.

[9]  Aboul Ella Hassanien,et al.  Fuzzy and hard clustering analysis for thyroid disease , 2013, Comput. Methods Programs Biomed..

[10]  Ethem Alpaydin,et al.  Constructive feedforward ART clustering networks. I , 2002, IEEE Trans. Neural Networks.

[11]  Zain Anwar Ali,et al.  Hybrid Anomaly Detection by Using Clustering for Wireless Sensor Network , 2019, Wirel. Pers. Commun..

[12]  Jirí Andel,et al.  Estimating the dimension of a linear model , 1981, Kybernetika.

[13]  C. C. Palerm,et al.  Closed-Loop Insulin Delivery Using a Subcutaneous Glucose Sensor and Intraperitoneal Insulin Delivery , 2009, Diabetes Care.

[14]  Robert A. Lordo,et al.  Learning from Data: Concepts, Theory, and Methods , 2001, Technometrics.

[15]  T. Soni Madhulatha,et al.  An Overview on Clustering Methods , 2012, ArXiv.

[16]  Ali Cinar,et al.  Hybrid Online Sensor Error Detection and Functional Redundancy for Artificial Pancreas Control Systems , 2016 .

[17]  Ali Cinar,et al.  Hybrid online sensor error detection and functional redundancy for systems with time-varying parameters. , 2017, Journal of process control.

[18]  A. M. Kalpana,et al.  Outlier analysis and Detection using K-medoids with support vector machine , 2016, 2016 International Conference on Computer Communication and Informatics (ICCCI).

[19]  Niels Kjølstad Poulsen,et al.  Fault and meal detection by redundant continuous glucose monitors and the unscented Kalman filter , 2017, Biomed. Signal Process. Control..

[20]  Janet M. Allen,et al.  Manual closed-loop insulin delivery in children and adolescents with type 1 diabetes: a phase 2 randomised crossover trial , 2010, The Lancet.

[21]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part I: Quantitative model-based methods , 2003, Comput. Chem. Eng..

[22]  Josep Vehí,et al.  Principal component analysis in combination with case-based reasoning for detecting therapeutically correct and incorrect measurements in continuous glucose monitoring systems , 2013, Biomed. Signal Process. Control..

[23]  Huang Chuanhe,et al.  Anomaly detection using Support Vector Machine classification with k-Medoids clustering , 2012, 2012 Third Asian Himalayas International Conference on Internet.

[24]  Josep Vehí,et al.  Using Support Vector Machines to Detect Therapeutically Incorrect Measurements by the MiniMed CGMS® , 2008, Journal of diabetes science and technology.

[25]  Ruizhe Ma,et al.  Distance and Density Clustering for Time Series Data , 2017, 2017 IEEE International Conference on Data Mining Workshops (ICDMW).

[26]  Chunhui Zhao,et al.  An effective fault detection method with FDA classifier and global model for continuous glucose monitor (CGM) , 2017, 2017 36th Chinese Control Conference (CCC).

[27]  Ali Cinar,et al.  Real-Time Model-Based Fault Detection of Continuous Glucose Sensor Measurements , 2017, IEEE Transactions on Biomedical Engineering.

[28]  Yao Wu,et al.  Ocean Data Anomaly Detection Algorithm Based on Improved k-medoids , 2019, 2019 Eleventh International Conference on Advanced Computational Intelligence (ICACI).

[29]  Chunhui Zhao,et al.  Statistical analysis based online sensor failure detection for continuous glucose monitoring in type I diabetes , 2015 .

[30]  Quan Shen,et al.  Online dropout detection in subcutaneously implanted continuous glucose monitoring , 2010, Proceedings of the 2010 American Control Conference.

[31]  Eyal Dassau,et al.  In silico evaluation platform for artificial pancreatic beta-cell development--a dynamic simulator for closed-loop control with hardware-in-the-loop. , 2009, Diabetes technology & therapeutics.

[32]  Giovanni Sparacino,et al.  An Online Failure Detection Method of the Glucose Sensor-Insulin Pump System: Improved Overnight Safety of Type-1 Diabetic Subjects , 2013, IEEE Transactions on Biomedical Engineering.

[33]  G. Freckmann,et al.  Accuracy of BG Meters and CGM Systems: Possible Influence Factors for the Glucose Prediction Based on Tissue Glucose Concentrations , 2016 .

[34]  K. Turksoy,et al.  Multivariable adaptive closed-loop control of an artificial pancreas without meal and activity announcement. , 2013, Diabetes technology & therapeutics.

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

[36]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part III: Process history based methods , 2003, Comput. Chem. Eng..

[37]  Jianhua Zhao,et al.  Mixture model selection via hierarchical BIC , 2015, Comput. Stat. Data Anal..

[38]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[39]  J Geoffrey Chase,et al.  Using Stochastic modelling to identify unusual continuous glucose monitor measurements and behaviour, in newborn infants , 2012, Biomedical engineering online.

[40]  B Wayne Bequette,et al.  Fault Detection and Safety in Closed-Loop Artificial Pancreas Systems , 2014, Journal of diabetes science and technology.

[41]  B. Wayne Bequette,et al.  Detecting sensor and insulin infusion set anomalies in an artificial pancreas , 2013, 2013 American Control Conference.

[42]  Ali Cinar,et al.  Multivariate Statistical Monitoring of Sensor Faults of A Multivariable Artificial Pancreas , 2017 .

[43]  Zuherman Rustam,et al.  Fuzzy Kernel k-Medoids algorithm for anomaly detection problems , 2017 .

[44]  Hae-Sang Park,et al.  A simple and fast algorithm for K-medoids clustering , 2009, Expert Syst. Appl..

[45]  Cun Ji,et al.  ADARC: An anomaly detection algorithm based on relative outlier distance and biseries correlation , 2020, Softw. Pract. Exp..

[46]  Lennart Ljung,et al.  System Identification: Theory for the User , 1987 .

[47]  Robert G. Sutherlin,et al.  A Bihormonal Closed-Loop Artificial Pancreas for Type 1 Diabetes , 2010, Science Translational Medicine.

[48]  Lior Rokach,et al.  A survey of Clustering Algorithms , 2010, Data Mining and Knowledge Discovery Handbook.

[49]  Ali Cinar,et al.  Monitoring and fault detection of continuous glucose sensor measurements , 2015, 2015 American Control Conference (ACC).

[50]  A. Çinar.  Multivariable Adaptive Artificial Pancreas System in Type 1 Diabetes , 2017, Current Diabetes Reports.

[51]  Ahmad Haidar,et al.  Closed-Loop Basal Insulin Delivery Over 36 Hours in Adolescents With Type 1 Diabetes , 2013, Diabetes Care.

[52]  Philip S. Yu,et al.  Top 10 algorithms in data mining , 2007, Knowledge and Information Systems.

[53]  P. Rousseeuw.  Silhouettes: a graphical aid to the interpretation and validation of cluster analysis , 1987 .