Healthcare Conference 2013 : Toekomstvisie op ICT in de gezondheidszorg - prof. dr. Bart De Moor

• 1. Toekomstvisie op ICT in de gezondheidszorg Prof. Dr. Bart De Moor Care for Innovation Healthcare Conference 12 November 2013, Affligemwww.esat.kuleuven.be/stadiuswww.kuleuven.be/iminds/fh

2. Contents Disruptive science and technology 3P/4P Medicine Preventive, Predictive, Personalized, Participatory Patients, Professionals, Policy Makers Exemplary cases Future Trends 3. www: max 19 clicks ! 4. 1 million = 1 000 000 1 billion = 1 000 000 000 1 trillion = 1 000 000 000 000 1 quadrillion = 1 000 000 000 000 0001 kB = 1 000 1 MB = 1 000 000 1 GB = 1 000 000 000 1 TB = 1 000 000 000 000 1 PB = 1 000 000 000 000 0001 TB = large university library = 212 DVD discs = 1430 CDs = 3 year music in CD quality 5. Connectivity We are alwaysCONNECTED and FAST!Throughput NOT sufficient for moving around Big Data 5 6. Tsunami of data from progress in technologyGS-FLX Roche Applied Science 454Sequencers Computer TomographyMagnetic resonanceMicroarrays (DNA chips)ACACATTAAATCTTATATGC TAAAACTAGGTCTCGTTTTA GGGATGTTTATAACCATCTT TGAGATTATTGATGCATGGT TATTGGTTAGAAAAAATATA CGCTTGTTTTTCTTTCCTAG GTTGATTGACTCATACATGT GTTTCATTGAGGAAGGAAC TTAACAAAACTGCACTTTTT TCAACGTCACAGCTACTTTA AAAGTGATCAAAGTATATCA AGAAAGCTTAATATAAAGAC ATTTGTTTCAAGGTTTCGTA AGTGCACAATATCAAGAAG ACAAAAATGACTAATTTTGT TTTCAGGAAGCATATATATT ACACGAACACAAATCTATTT TTGTAATCAACACCGACCAT GGTTCGATTACACACATTAA ATCTTATATGCTAAAACTAG GTCTCGTTTTAGGGATGTTT ATAACCATCTTTGAGATTAT TGATGCATGGTTATTGGTTA GAAAAAATATACGCTTGTTT TTCTTTCCTAGGTTGATTGAMass spectrometry 7. Genome data Human genome project Initial draft: June 2000 Final draft: April 2003 13 year project $300 million value with 2002 technology Personal genome Year June 1, 2007 Genome of James Watson, co-discoverer of DNA double helix, is sequenced $1.000.000 Two months 1000-genome Expected 2012-20201,00E+11 1,00E+10 1,00E+09 1,00E+08 1,00E+07 1,00E+06 1,00E+05 1,00E+04 1,00E+03 1,00E+02 1,00E+01 1,00E+00 1,00E-01 1,00E-02 1,00E-03 1,00E-04 1,00E-05 1,00E-06 1,00E-07Cost per base pair Genome cost199019952000200220052007Cost per base pair20102015Genome cost1990103E+10199513.000.000.00020000.2600.000.00020020.09270.000.00020050.0390.000.00020070.0003333331.000.00020103.33333E-061000020150.00000013007 8. index of 20 million Biomedical PubMed records1 slice mouse brain MSI at 10 m resolution 81 GigaByteraw NGS data of 1 full genomesequencing all newborns by 2020 (125k births / year) 125 PetaByte / year1 TeraByte23 GigaByte1 kB = 1000 1 MB = 1 000 000 1 GB = 1 000 000 000 1 TB = 1 000 000 000 000 1 PB = 1 000 000 000 000 000 1 small animal image 1 CDROM 750 MegaByt e1 GigaBytePACS UZ Leuven 1,6 PetaByteGenomics core HiSeq 2000 full speed exome sequencing 1 TeraByte / week8 9. Contents Disruptive science and technology 3P/4P Medicine Preventive, Predictive, Personalized, Participatory Patients, Professionals, Policy Makers Exemplary cases Future Trends 10. ICT tools for decision support ICT to support 3P decision makers in health care Professionals (clinicians, biomedical researchers ) exploit data for more effective medicine Patient patient-centered care for empowered patients Policy Maker (mutualities, health insurers, social security, governments, hospital management ) Smart, data-driven, evidence-based health care system policies10 11. ICT contributes to P4 medicine Personalized: "customized" diagnosis and treatment, right drug for the right patient at the right time Preventive: prevention and early diagnosis Predictive: determine risk profiles, predict progression and outcome Participatory: correct and complete information for the patient to participate in the decision process, selfmanagement11 12. Rationales for ICT & Health -Improve quality performance of health decision/diagnosis systems -Support individual medical doctor -Avoid/decrease number of medicial errors -Web portal for Evidence Based Medicine -Information sharing among doctors -avoid/monitor patient (s)hopping behavior -Global Medical File per patient -Interoperability-Deal with empowerment of the patient: Patient-centric health care -Medical care in 4P: personalized, preventive, predictive, participatory -Increasing trend for customizedpersonalized medicine -Improve transparancy and consistency -Deal/cope with professional (chronical) patients (heart, diabetes, cancer,) -Improve patient mobility -Cost effectiveness of the health care system -Ageing population: -EU 2050: 65+ +70%; 80+ +180% -Monitor overconsumption -Improve transparancy -Detect abnormalities in diagnosis/therapy/ -Cope with tsunami of available information and data (clinical, population, .) 13. Contents Disruptive science and technology 3P/4P Medicine Preventive, Predictive, Personalized, Participatory Patients, Professionals, Policy Makers Exemplary cases Future Trends 14. NXT_SLEEP: Next generation sleep monitoring platform14 15. FallRisk: Intelligent services in view of prevention, risk evaluation and detection of fall incidents15 16. Blood Glucose Control @ ICU 10 mio adult ICU patients / year (EU + US) (1-2 bio $ market) Intensive Care Unit UZ Leuven studies: Tight Glycemic Control (TGC) in intensive care unit lowers mortality LOGIC-Insulin: semi-automatic control system that advises nurse on insulin dosage and blood sampling interval aiming at TGC and avoiding hypoglycemia clinical trial: compared with expert nurses, LOGIC-Insulin showed improved efficacy of TGC without increasing rate of hypoglycemia multicentre clinical trial Q3 2013; spin-off in preparation16 17. IOTA app: ovarian tumour malignancy: population based / standardizedIOTA Models:Standardize ultrasonographic endometrial tumor analysis models giving an indication of the probability of malignancy of an ovarian tumour based on 6 to 12 observed parameters IOTA app available in iTunes app store and on http://homes.esat.kuleuven.be/~sistawww/biomed/iota/ 17 18. IOTA app: Clinical Decision Support goal = making it easier to diagnose ovarian cancer help clinicians take decisions on diagnosis, prognosis and therapy response make use of patient data nd population information (patient biobank & database, literature ) build mathematical model on data and use this model to predict patient outcome: formalized way of analyzing biomedical data instead of interpretation of clinical parameters based on a clinicians expert knowledge model method can be varied: logistic regression, artifical neural networks, Support Vector Machines (SVM), Bayesian networks Patient history Tumor characteristicsClinicianUltrasound characteristics Tumor markersDiagnosisPrognosisModelResponse to therapy 19. IOTA performance comparison Performance of an expertPerformance of IOTA modelsPerformance of non-expertsPerformance of old models 20. Microarrays DNA-chips Test Ref.High Low LowHighHigh HighLowLow 21. Genetic classification of leukemia 12600 genes / 72 patients / 3 types of leukemia (AML/ ALL/MLL) Data matrixFind the patternHidden patternPattern validation 22. 18 / 21 patients with 87 gene fingerprint for AML19 / 25 patients with 80 gene fingerprint for ALL14 / 17 patients with 62 gene fingerprint for MLL 23. ALL/AML/MLL biomarkers PCA Armstrong SA et al. Nat Genet. 2002 Jan;30(1):41-7.12 600 genes 72 patients: - 28 Acute Lymphoblastic Leukemia (ALL) - 24 Acute Myeloid Leukemia (AML) - 20 Mixed Linkage Leukemia (MLL) 3 patients for each class used as test set 24. Aerts et al, Nature Biotechnology, 2006 24 25. 25 26. 26 27. Variant prioritization by genomic data fusion: including disease phenotypesFigure Classification scenarios. Receiver-operator (A and C) and Precision-Recall (B and D) curves for five classifiers tested against test sets comparing disease-causing variants to either rare non-disease-causing variants (A and B) or common polymorphisms (C and D).based on Random Forests 20x more accurate than state of the artSifrim, Popovic et al, Nature Methods, 2013d 28. Mass Spectral Imaging Mass spectral imaging allows Screening and mapping of hundreds of biomolecules in a complex organic tissue section Does not require any prior information nor labeling Visualization of the spatial distribution of these moleculesComputation complexity The structure of the generated data is complex Large size and high dimensionalityin collaboration with SyBioMa, KU Leuven Facility for Systems Biology based Mass Spectrometry, MIRC KU Leuven & Caprioli Lab (Vanderbilt University, VS) 29. Contents Disruptive science and technology 3P/4P Medicine Preventive, Predictive, Personalized, Participatory Patients, Professionals, Policy Makers Exemplary cases Future Trends 30. Moores law: Computing power doubles every 18 months (56 % growth rate) Understand ?6 10 9Operations/second5 10 9LUI4 10 93D games 3 1092 10 91 10 9Bookkeeping 0 19751980Audio19851990Video1995Year200020052010 31. Body area networksEEGPOTSVisionwww Networkpositioning glucose CellularTransducer Nodes Hearing ECG Blood pressureDNA proteinImplantsWLANPersonal Assistant 31 32. 1953 Crick and Watson: DNA Double Helix2000 Venter, Collins: Humane genome 33. E D. Green et al. Nature 470, 204-213 (2011) doi:10.1038/nature09764Towards effective genomic medicine 34. building on expertise in big data & machine learning BENCH is used in ca. 50 accredited genetic labs worldwide Heverlee + US officeroutine diagnosticstools for non-IT usersfirst diagnostics grade solution for Next Generation Sequencing data based diagnostics in the worldoriginating from research in rare genetic disorders in close collaboration with Centre Human Genetics UZ Leuven 34 35. Systems biology: reverse engineeringhigh throughput datagenometranscriptomeproteomemetabolomeinteractome 35 36. Zebrafish Neural Circuits understand dynamical behaviour of neural circuits in Zebrafish through system identification uy = f (u )yInputRAW STIMULUSOutputBRAINModelIdentification AlgorithmREACTION-in collaboration with NERF 37. Signal Processing and Classification for Magnetic Resonance Spectroscopy Applied to Brain Tumor Diagnosis Single-voxel MRSMRS quantificationNAAMyoCrPChoGluLacLip1Lip2AlaGlcMetabolite concentrationsTauBiomarker of diseaseMetabolite maps Multi-voxel MRSMRS quantification using spatial information 38. Seizure detection incorporating structural information from the multichannel EEGRepresentation of EEG data in higher order arrays:features channelsfrequencychannelsfeatureschannelschannelstimeClassifier matrixGoal: Exploit the inherent structure of the EEG signals using novel matrix and tensor-based machine learning solutions: Nuclear norm regularization Tensorial kernels 39. Collaboration IMEC/NERF: neuroprobes MULTI-FUNCTIONAL INTEGRATED NEUROPROBES ADVANCED & AUTOMATED MULTICHANNEL SIGNAL PROCESSING... INTEGRATED END-USER FEEDBACK 40. Synthetic biology: design new functional life Bacterium detecting cancer cells Output Input Reset MemoryFilter Cell Death InverTimerhttp://2013.igem.org/Team:KU_Leuven 41. Obamahttp://www.whitehouse.gov/blog/09/04/27/The-Necessity-of-Science/ The Recovery Act will support the long overdue step of computerizing America's medical records, to reduce the duplication, waste and errors that cost billions of dollars and thousands of lives. But it's important to note, these records also hold the potential of offering patients the chance to be more active participants in the prevention and treatment of their diseases. We must maintain patient control over these records and respect their privacy. At the same time, we have the opportunity to offer billions and billions of anonymous data points to medical researchers who may find in this information evidence that can help us better understand disease. . Because of recent progress - not just in biology, genetics and medicine, but also in physics, chemistry, computer science, and engineering - we have the potential to make enormous progress against diseases in the coming decades. .41 42. Decision Support Analytics for Professionals, Patients & Policy [email protected]/stadiuswww.kuleuven.be/iminds/fh