Wearable Monitoring and Interpretable Machine Learning Can Objectively Track Progression in Patients during Cardiac Rehabilitation

Cardiovascular diseases (CVD) are often characterized by their multifactorial complexity. This makes remote monitoring and ambulatory cardiac rehabilitation (CR) therapy challenging. Current wearable multimodal devices enable remote monitoring. Machine learning (ML) and artificial intelligence (AI) can help in tackling multifaceted datasets. However, for clinical acceptance, easy interpretability of the AI models is crucial. The goal of the present study was to investigate whether a multi-parameter sensor could be used during a standardized activity test to interpret functional capacity in the longitudinal follow-up of CR patients. A total of 129 patients were followed for 3 months during CR using 6-min walking tests (6MWT) equipped with a wearable ECG and accelerometer device. Functional capacity was assessed based on 6MWT distance (6MWD). Linear and nonlinear interpretable models were explored to predict 6MWD. The t-distributed stochastic neighboring embedding (t-SNE) technique was exploited to embed and visualize high dimensional data. The performance of support vector machine (SVM) models, combining different features and using different kernel types, to predict functional capacity was evaluated. The SVM model, using chronotropic response and effort as input features, showed a mean absolute error of 42.8 m (±36.8 m). The 3D-maps derived using the t-SNE technique visualized the relationship between sensor-derived biomarkers and functional capacity, which enables tracking of the evolution of patients throughout the CR program. The current study showed that wearable monitoring combined with interpretable ML can objectively track clinical progression in a CR population. These results pave the road towards ambulatory CR.

[1]  M. Bahls,et al.  Towards a personalised approach in exercise-based cardiovascular rehabilitation: How can translational research help? A ‘call to action’ from the Section on Secondary Prevention and Cardiac Rehabilitation of the European Association of Preventive Cardiology , 2019, European journal of preventive cardiology.

[2]  A. Coats,et al.  Chronotropic incompetence in chronic heart failure. , 1995, International journal of cardiology.

[3]  Gunnar Hartvigsen,et al.  Using Fitness Trackers and Smartwatches to Measure Physical Activity in Research: Analysis of Consumer Wrist-Worn Wearables , 2018, Journal of medical Internet research.

[4]  Michael Weis,et al.  Accelerometer-based quantification of 6-minute walk test performance in patients with chronic heart failure: applicability in telemedicine. , 2009, Journal of cardiac failure.

[5]  V. Kecman,et al.  Iterative Single Data Algorithm for Training Kernel Machines from Huge Data Sets: Theory and Performance , 2005 .

[6]  Bor-Shyh Lin,et al.  Wearable Cardiopulmonary Function Evaluation System for Six-Minute Walking Test , 2019, Sensors.

[7]  P. Ponikowski,et al.  [2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure]. , 2016, Kardiologia polska.

[8]  N. Pogosova,et al.  [Secondary prevention in the clinical management of patients with cardiovascular diseases. Core components, standards and outcome measures for referral and delivery]. , 2014, Kardiologiia.

[9]  Biykem Bozkurt,et al.  2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. , 2013, Circulation.

[10]  Sabine Van Huffel,et al.  Robust artefact detection in long-term ECG recordings based on autocorrelation function similarity and percentile analysis , 2012, 2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[11]  Volkmar Falk,et al.  2016 ESC Guidelines for the Diagnosis and Treatment of Acute and Chronic Heart Failure. , 2016, Revista espanola de cardiologia.

[12]  J. Cafazzo,et al.  Use of Free-Living Step Count Monitoring for Heart Failure Functional Classification: Validation Study , 2019, JMIR cardio.

[13]  Robert Wang,et al.  Accuracy of wearable heart rate monitors in cardiac rehabilitation. , 2019, Cardiovascular diagnosis and therapy.

[14]  Theresa M. Beckie,et al.  2018 ACC/AHA Clinical Performance and Quality Measures for Cardiac Rehabilitation: A Report of the American College of Cardiology/American Heart Association Task Force on Performance Measures. , 2018, Journal of the American College of Cardiology.

[15]  Assessing the Use of Wrist-Worn Devices in Patients With Heart Failure: Feasibility Study , 2017, JMIR cardio.

[16]  Tom Dhaene,et al.  Systematic Comparison of Respiratory Signals for the Automated Detection of Sleep Apnea , 2018, 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC).

[17]  L. Neubeck,et al.  Validation of Fitbit-Flex as a measure of free-living physical activity in a community-based phase III cardiac rehabilitation population , 2016, European journal of preventive cardiology.

[18]  Mohammad Hossein Khosravi,et al.  Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980–2017: a systematic analysis for the Global Burden of Disease Study 2017 , 2018, Lancet.

[19]  Francisco B. Ortega,et al.  Accelerometer Data Collection and Processing Criteria to Assess Physical Activity and Other Outcomes: A Systematic Review and Practical Considerations , 2017, Sports Medicine.

[20]  Julien Penders,et al.  Cardiorespiratory fitness estimation in free-living using wearable sensors , 2016, Artif. Intell. Medicine.

[21]  D. Lipkin,et al.  Cardiac rehabilitation. , 1993, BMJ.

[22]  C. Craig,et al.  Physical activity of Canadian adults: accelerometer results from the 2007 to 2009 Canadian Health Measures Survey. , 2011, Health reports.

[23]  Kipp W. Johnson,et al.  Machine learning in cardiovascular medicine: are we there yet? , 2018, Heart.

[24]  P. Sinnaeve,et al.  Assessment of Physical Activity by Wearable Technology During Rehabilitation After Cardiac Surgery: Explorative Prospective Monocentric Observational Cohort Study , 2019, JMIR mHealth and uHealth.

[25]  R. McKelvie,et al.  Characterization of subgroups of heart failure patients with preserved ejection fraction with possible implications for prognosis and treatment response , 2015, European journal of heart failure.

[26]  Michele Angelaccio,et al.  Remote Patient Monitoring via Non-Invasive Digital Technologies: A Systematic Review , 2017, Telemedicine journal and e-health : the official journal of the American Telemedicine Association.

[27]  Theresa M. Beckie,et al.  Home-Based Cardiac Rehabilitation: A SCIENTIFIC STATEMENT FROM THE AMERICAN ASSOCIATION OF CARDIOVASCULAR AND PULMONARY REHABILITATION, THE AMERICAN HEART ASSOCIATION, AND THE AMERICAN COLLEGE OF CARDIOLOGY. , 2019, Journal of cardiopulmonary rehabilitation and prevention.

[28]  Sandra Prescher,et al.  Tele-accelerometry as a novel technique for assessing functional status in patients with heart failure: feasibility, reliability and patient safety. , 2013, International Journal of Cardiology.

[29]  Luo Xiao,et al.  An Activity Index for Raw Accelerometry Data and Its Comparison with Other Activity Metrics , 2016, PloS one.

[30]  Bruce R. Schatz,et al.  A Natural Walking Monitor for Pulmonary Patients Using Mobile Phones , 2015, IEEE Journal of Biomedical and Health Informatics.

[31]  I. Frederix,et al.  Who needs secondary prevention? , 2017, European journal of preventive cardiology.

[32]  Geoffrey E. Hinton,et al.  Visualizing Data using t-SNE , 2008 .

[33]  Sabine Van Huffel,et al.  R-DECO: an open-source Matlab based graphical user interface for the detection and correction of R-peaks , 2019, PeerJ. Computer science.

[34]  ATS statement: guidelines for the six-minute walk test. , 2002, American journal of respiratory and critical care medicine.

[35]  Prashant Warier,et al.  Machine Learning Methods Improve Prognostication, Identify Clinically Distinct Phenotypes, and Detect Heterogeneity in Response to Therapy in a Large Cohort of Heart Failure Patients , 2018, Journal of the American Heart Association.

[36]  Lionel Tarassenko,et al.  The Mobile-Based 6-Minute Walk Test: Usability Study and Algorithm Development and Validation , 2020, JMIR mHealth and uHealth.

[37]  Tapas Mondal,et al.  Wearable Sensors for Remote Health Monitoring , 2017, Sensors.

[38]  R. Shephard Physical activity of Canadian adults: Accelerometer results from the 2007 to 2009 Canadian Health Measures Survey , 2011 .

[39]  Ruairi O'Driscoll,et al.  How well do activity monitors estimate energy expenditure? A systematic review and meta-analysis of the validity of current technologies , 2020, British Journal of Sports Medicine.

[40]  S. Steinhubl,et al.  Connected Health Technology for Cardiovascular Disease Prevention and Management , 2019, Current Treatment Options in Cardiovascular Medicine.

[41]  J. Imich Home-based cardiac rehabilitation. , 1997, Nursing times.

[42]  Hugo Saner,et al.  Chronotropic incompetence predicts impaired response to exercise training in heart failure patients with sinus rhythm , 2013, European journal of preventive cardiology.

[43]  A. Holland,et al.  The six‐minute walk test: a useful metric for the cardiopulmonary patient , 2009, Internal medicine journal.

[44]  Sabine Van Huffel,et al.  R-DECO: an open-source Matlab based graphical user interface for the detection and correction of R-peaks , 2019, bioRxiv.