Realizing real time monitoring and controlling bio-signals Daniel Berckmans M3 BIORES Katholieke Universiteit Leuven Belgium Philips Eindhoven April 11 2008
K.U.Leuven DIVISION M3-BIORES Measure, Model and Manage Bioresponses Head: Prof. Daniel Berckmans Coordinators Imperfectly mixed fluids Bioresponses 1 Process output & environment Bioresponses 2 Ir. SEZIN EREN OZCAN Prof. JEAN-MARIE AERTS Dr. Claudia Bahr Dr. STIJN QUANTEN Dr. TADIWOS ZERIHUN DESTA Ir. KRISTIEN VAN LOON Prof. ERIK VRANKEN Ir. PIETER SCHIEPERS Ir. VO TAN THANH Ir. GUIDO DE BRUYNE Ir. OZLEM CANGAR Ir. DRAGANA MILJKOVIC Ir. NILA ALBAN Ir. JORIS LEFEVER Ir. SONG XIANGYU Ir. X Ir. SEBASTIAN DE BOODT Ir. TOON LEROY Ir. X Ir. X Ing. KAREL DE BODT Ir. X Ir. X Ir. JAN DEKOCK Ir. SARA FERRARI Ir. X Secretary ANN VAN GINDERACHTER Sound analysis and bioresponses Dr. VASILEIOS EXADAKTYLOS Ir. MITCHELL SILVA BioRICS M3-BIORES, NV, K.U.LEUVEN Belgium Technical assistance Ing. JEAN-LOUIS LEMAIRE LUDO HAPPAERTS PRIVATE COMPANIES BioRICS NV Ir. KIM BATSELIER Ir. FREDERIK JANSEN Ing. MIHAJLO MILUTINOVIC Ir. X Ir. X 2
BioRICS n.v. - K.U. Leuven, LRD ( o 1972): 56 spin-offs e.g. LMS, Ubizin, Thromb-X,... - 2006: 3
Problem Livestock farming in the past Farmer had the time to use audio-visual scoring 4
Today however Low price of product High number of animals More welfare problems Less available time per animal 5
258 million ton meat/year 7 billion pigs 40 billion chicken Market numbers 6
A living organism: Complex 7
A living organism: Complex Individual Heartbeat (bpm) Time (s) 8
Identical Individually different 9
A living organism: Complex Individual Time-Varying Example: Heat production of broiler chickens 17 5 days old 13 30 days old HEAT PRODUCTION (W/KG) 16 15 14 13 12 11 0 1 2 3 4 5 MEASURED MODELLED (1ST ORDER) MODELLED (2ND ORDER) TIME (HOURS) HEAT PRODUCTION (W/KG) Different model order 12 11 10 9 8 7 0 1 2 3 4 5 MEASURED MODELLED (1ST ORDER) TIME (HOURS) 10
A living organism: Complex Individual Time-Varying Dynamic Living organism = CITD - system Measure Model Manage In an on-line way M3-BIORES 11
What happens today in other process control 12
General methodology: Conditions for process control Position Steering wheel Process Direction MEASURE 2 PREDICTION 3 MEASURE 1 FEEDBACK Desired direction MODEL-BASED CONTROLLER 13
Control of Biological Responses Process MICRO- ENVIRONMENT DYNAMIC BIORESPONSE 2 PREDICTION 1 FEEDBACK 3 MODEL BASED CONTROLLER DESIRED BIORESPONSE 14
Examples 15
Control of the crawl trajectory of larvae of Calliphora vicina 16
Control of the crawl trajectory of larvae of Calliphora vicina Light intensity crawl trajectory 17
Response of crawl direction to variations in ligth Light intensity u(t) t crawl trajectory y(t) t MEASURE MEASURE 18
Responses of crawl direction to steps in light direction 360.00 315.00 270.00 Crawl direction ( ) 225.00 180.00 135.00 90.00 45.00 270.00 180.00 12.55 315.00 270.00 9.72 45.00 90.00 13.59 90.00 180.00 15.77 225.00 90.00 12.77 135.00 270.00 9.60 0.00 0.00 15.60 180.00 0.00 15.61 0.00-45.00 0 3 6 9 12 15 18 21 24 27 30 33 36 Time 19
Active control of crawl direction of larvae Light intensity crawl trajectory MEASURE MEASURE PREDICTION FEEDBACK MODEL-BASED CONTROLLER 20
Active control of crawl direction of 6 larvae 21
Control of the growth trajectory of chickens 22
Active on-line growth trajectory control Chickens > 2000 g Weight Problems! Leg disorders Ascites Sudden death syndrome 40 g Day 42 Day 1 23
Possible solution: Growth restriction + compensatory growth = growth control Weight Growth restriction Compensatory growth Day 1 Day 42 24
Control diagram Temperature Light schedule Feed quantity Feed quality Weight trajectory t Prediction Feedback Model Model based controller 25
EXP2001-3 - afdeling2 2500 Gewicht (gram) Weight (grams) 2000 1500 1000 500 0 0 10 20 30 40 Tijd (dagen) Time (days) Controlled Gestuurd groeitraject growth trajectory Reference Referentie growth groeitraject trajectory Growth Groeitraject trajectory controle of control groep group = 3.7% - 6.0% PLF control results in reduction of mortality (12 exp, 2900 animals each) of 4% 26
Optimisation of physical training of race horses 27
Optimisation of physical training of race horses Running speed Heart rate v VO 2 Heart rate Lactate conc. 28
Sensors Polar equine HR monitor Garmin forerunner GPS 29
Optimisation of physical training of race horses Velocity and Heart rate - Kyrielle (Bert) 01-01 30 140 velocity (km/u) 25 20 15 10 Velocity Heart rate 120 100 80 60 40 heart rate (bpm) 5 20 0 0 0:00:00 0:05:00 0:10:00 0:15:00 0:20:00 0:25:00 time (u:mm:ss) 30
Optimisation of physical training of race horses 120 110 heart rate (bpm 100 90 80 70 60 50 40 Controlled heart rate Target heart rate 0:00:00 0:10:00 0:20:00 0:30:00 0:40:00 0:50:00 1:00:00 time(h:mm:ss) 31
Optimisation of physical training of race horses Time-constant of HR over time 70 60 TC 50 TC (sec) 40 30 20 10 0 Week 0 Week 1 Week 2 Week 3 Week 4 Time (date) 32
Model based monitoring of physical & mental performance (AC Milan) 33
Process Physical training Physical performance Mental training CITD CITD Mental performance Performance Total performance = a. Mental performance + b. Physical performance 34
Process Training exercises Performance MEASURE (Inmotio) Body MEASURE (Inmotio + Hosand) 35
Mental monitoring Heart Rate Physical Activity Mental algorithm On-line mental monitor Time??? Reference mental status 36
Reference mental status MINDROOM 37
Demo: mental monitoring 38
Mindroom data (1133 and 1156) 1.5 1 0.5 0-0.5-1 -1.5 39 27/03/2006 31/03/2006 3/4/2006 7/4/2006 13/4/2006 19/4/2006 24/4/2006 27/4/2006 1/5/2006 2/5/2006 3/5/2006 4/5/2006 10/5/2006 27/03/2006 31/03/2006 7/4/2006 11/4/2006 Mindroom vs. Mental algorithm 1133 1156 Mental status [pos, neg or neutral] Date (ddmmyy) MINDROOM
Mindroom vs. Mental algorithm Mental algo vs Mindroom 1.5 1 1133 1156 0.5 0-0.5-1 -1.5 40 11/4/2006 Mental status [pos, neg or neutral] 27/03/2006 31/03/2006 3/4/2006 7/4/2006 13/4/2006 19/4/2006 24/4/2006 27/4/2006 1/5/2006 2/5/2006 3/5/2006 4/5/2006 10/5/2006 27/03/2006 31/03/2006 7/4/2006 Date (ddmmyy) MINDROOM MENTAL ALGO
Real-time process control Physical training BioRICS Physical performance Mental training CITD CITD BioRICS Mental performance Performance Total performance = a. Mental performance + b. Physical performance 41
Application Examples 42
Stressmonitoring USB transceiver BioRICS server Accelerometer ECG Mobile phone 43
Monitoring and control of physical performance USB transceiver BioRICS server Accelerometer ECG Mobile phone 44
Weight Control Accelerometer ECG Mobile phone USB transceiver BioRICS server 45
Driver Sleepiness Pedal + steer input USB transceiver BioRICS server Temperature sensor ECG 46
General hardware scheme Sensor Modules ECG module Temp. module Accelerometer module1 Accelerometer module 3 Bidirectional Wireless link 1 2m Central Module PDA or Mobile Bidirectional Wireless link 2 www.biorics.com Central Module : Transceiver connected to PDA or mobile (mini-usb) Responsible for synchronization of the signals from the different sensor modules 47
Specifications ECG sensor module : sampling freq : 300 Hz Temperature module : sampling freq : 1/60Hz (1 sample/minute) Accelerometer modules : sampling freq : 200 Hz g range : 6g ; 10g ; 25g Bidirectional Wireless link 1 : wireless range : 2m Strongly dependent on application field low power (e.g. recharging after 1 week) Central Module : USB transceiver Synchronization of the different signals form the sensor modules Data transmission via mobile network to BioRICS server 48
Complex Algorithms (high level language, e.g. Matlab) Translation Low level language (C) Compiler 49
Daphnia applications Mussels Chickens Tubifex 10 Products ESA Race horses Intensive Care Bicycle helmets Professional Cyclists 11 Patents 50
Thank you for your attention For more information you can always check our website: http://www.m3-biores.be http://www.biorics.com 51