Health-enabling technologies for pervasive health care: A pivotal - PowerPoint PPT Presentation

Health-enabling technologies for pervasive health care: A pivotal field for future medical informatics research and education? Reinhold Haux Matthias Gietzelt, Nils Hellrung, Wolfram Ludwig, Michael Marschollek, Bianying Song, Markus Wagner, Klaus-Hendrik Wolf Peter L. Reichertz Institute for Medical Informatics University of Braunschweig - Institute of Technology and Hannover Medical School Medical Informatics Europe 2009, Sarajevo, Bosnia and Herzegovina Peter L. Reichertz Institute for Medical Informatics

structure • background: the demographic change • health-enabling technologies and pervasive health care • examples • health-enabling technologies and medical informatics Peter L. Reichertz Institute for Medical Informatics

structure • on the Peter L. Reichertz Institute • background: the demographic change • health-enabling technologies and pervasive health care • examples • health-enabling technologies and medical informatics Peter L. Reichertz Institute for Medical Informatics

references Bardram JE. Pervasive healthcare as a scientific discipline. • Methods Inf Med. 2008; 47: 178-85. Haux R et al. Health-enabling technologies for pervasive health care. • Inform Health Soc Care. 2008; 33: 77-89. Koch S et al. On Health-enabling and ambient-assistive technologies. • Methods Inf Med. 2009; 48: 29-37. Saranummi N. IT applications for pervasive, personal, and • personalized health. IEEE Trans Inf Technol Biomed. 2008; 12: 1-4. Saranummi N, Woctlar H, editors. Pervasive Healthcare. • Methods Inf Med. 2008; 47: 175-240. GAL: www.altersgerechte-lebenswelten.de • PLRI: www.plri.de • Peter L. Reichertz Institute for Medical Informatics

on the 
 Peter L. Reichertz Institute Peter L. Reichertz Institute for Medical Informatics

the Peter L. Reichertz Institute (PLRI) • since more than three decades medical informatics with Professor Peter L. Reichertz as pioneer • in 2007: University of Braunschweig - Institute of Technology (TU Braunschweig) and Hannover Medical School (MHH) unite their medical informatics institutes as a Prof. Reichertz joint institute, named Peter L. Reichertz 1930 - 1987 Institute for Medical Informatics with two locations Braunschweig and Hannover • aim: regional ‘center of excellence‘ • 2 locations, PLRI staff is member of both universities Peter L. Reichertz Institute for Medical Informatics

the PLRI • fields of research • health-enabling technologies • eLearning in medicine and dentistry • health information systems and management • medical imaging and visualization • education • medical informatics courses • medical informatics program (B.Sc, M.Sc., Ph.D) Peter L. Reichertz Institute for Medical Informatics

background: 
 the demographic change Peter L. Reichertz Institute for Medical Informatics

the demographic change adolescenting („aging“) societies „the number of persons aged 60 years or older is estimated to be 629 million in 2002 and is projected to grow to almost 2 billion by 2050, at which time the population of older persons will be larger than the population of children (0-14 years) for the first time in human history“ (UN Population Division) Peter L. Reichertz Institute for Medical Informatics

the demographic change Potential Support Ratio 1950-2050 World 12 9 4 2000 2050 1950 year source: UN, Population Division Peter L. Reichertz Institute for Medical Informatics

the demographic change Potential Support Ratio 1950-2050 World / Europe 12 9 8 4 5 2 2000 2050 1950 year source: UN, Population Division Peter L. Reichertz Institute for Medical Informatics

health-enabling technologies 
 and pervasive health care Peter L. Reichertz Institute for Medical Informatics

on health-enabling technologies (HET) • HET are information and communication technologies for creating sustainable conditions for self-sufficient and self-determined lifestyles • sensor-enhanced health information systems play a major role in this context, aiming to enable ambient-assisted living • by utilizing advanced HET, individual quality of life is intended to be enhanced, while sustaining the efficiency of health care Peter L. Reichertz Institute for Medical Informatics

on pervasive health care (from Jacob Bardram): acute  � continuous hospitalization  home & outpatient reactive  proactive & preventive IT  assistive technology centralizend  pervasive sampling  monitoring doctor-centric  patient-centric Bardram J. Pervasive Healthcare as Discipline Methods Inf Med. 2008; 47: 178-185. Peter L. Reichertz Institute for Medical Informatics

on HET & pervasive helth care: Niilo Saranummi’s 3‘P‘s • pervasive technologies shall enable semantically interoperable platforms to communicate and store health data and the use of health-enabling technologies • personal services using sensor technologies for continuously measuring health related data of an individual; to support her or him at specific health problems • personalized decision support adapted, ‘tuned’ to the individual’s norm, not to averages in populations Peter L. Reichertz Institute for Medical Informatics

on HET: financial considerations • example for possible cost savings through health-enabling technologies for Germany: • if elder citizens can stay for three month longer in their homes, we will save 315 million € / year Peter L. Reichertz Institute for Medical Informatics

on HET: opportunities • in our adolescenting/ageing societies, advanced HET may permit active, self-sufficient and autonomous lifestyles at high quality for increasing numbers of fellow citizens • sensor-enhanced health information systems for ambient-assisted living will be critical for early detection and prevention of diseases in the pre- clinical stage, as well as for alleviating chronic diseases Peter L. Reichertz Institute for Medical Informatics

the double circle Informatics for Health and Social Care. 2008; 33, 77 - 89. Peter L. Reichertz Institute for Medical Informatics

examples Peter L. Reichertz Institute for Medical Informatics

Demonstration • ECG and triaxial accelerometer • long-term monitoring of ECG/HRV with respect to • activity intensity • different activities of daily living (ADL) under real-life conditions  deduction: physiologic reaction of the cardio- vascular system to physical stress  long-term changes? Peter L. Reichertz Institute for Medical Informatics

Measuring movement - accelerometers • cheap, small, mobile/wearable standing up walking turning walking sitting down Peter L. Reichertz Institute for Medical Informatics

Long-term activity monitoring • N=1, young and healthy • Device: Sensewear Pro2 (multisensor) • duration: 6 months • data: 201.984 minutes, 100.087 annotated (=1,668 hrs.) • 28 activities, no preselection • activity classification with pattern recognition algorithms OneR, Naive Bayes, C4.5 Peter L. Reichertz Institute for Medical Informatics

Long-term activity monitoring - results activity classification accuracy for data sets classifier 9s 21s 60s OneR 78.0% 71.9% 69.2% Naive Bayes 81.1% 80.5% 80.6% C4.5 94.7% 92.4% 90.0% Marschollek et al., 2006 Peter L. Reichertz Institute for Medical Informatics

Activities class number of classification class number of classification instances accuracy instances accuracy gardening 1,973 85.4% sleeping while supine 286,732 99.3% standing still 1,073 85.3% working at the computer 155,763 95.6% telephoning 2,212 84.6% watching TV 73,958 95.0% eating 14,628 84.6% working at a constr. site 10,872 95.0% household cleaning 2,389 83.3% dancing 1,176 94.9% folding laundry 149 80.5% walking 24,824 94.3% meeting 22,727 80.3% sleeping while sitting 303 93.7% going to the bathroom 3,118 79.8% party 2,564 92.7% working at the office 8,998 79.6% driving a vehicle 624 90.4% sitting 7,666 75.5% watching movie at a cinema 1,581 90.3% waiting at traffic light 336 73.8% lying down 25,420 90.1% shopping 1,027 71.5% teaching 4,732 89.0% reading 2,535 69.1% packing and moving 1,544 88.4% traveling in a vehicle 1,546 64.0% Peter L. Reichertz Institute for Medical Informatics

Activity profiles of elderly persons • N=5, ∅ 67 years • Device: Sensewear Pro2 • data: 69.808 minutes • 17 activities (with >100min.) • activity classification with pattern recognition algorithm C4.5 • evaluation: • 10x 10-fold cross-validation ( intra individual) • „leave-one-out“ cross-validation ( inter individual) Peter L. Reichertz Institute for Medical Informatics

Activity profiles – results interindividual classification intraindividual classification person no. classification accuracy person no. classification accuracy 1 93.1% 1 32.8% 2 90.9% 2 52.3% 3 78.6% 3 27.8% 4 99.2% 4 87.9% 5 95.4% 5 67.8% mean 91.4% mean 53.7% Marschollek et al., 2007 SmarTel Best Paper Award Peter L. Reichertz Institute for Medical Informatics