Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Physiologic Monitoring in the Cloud
Carolyn McGregor AM, PhD Canada Research Chair Health Informatics Professor Faculty of Business and IT University of Ontario Institute of Technology Oshawa, Ontario, Canada
Professor Carolyn McGregor is the Canada Research Chair in Health Informatics at the University of Ontario Institute of Technology, Canada. Dr McGregor has led pioneering research in Big Data, analytics, event stream processing, temporal data stream data mining, business process modelling, patient journey modelling and cloud computing. In the 1990s she led two of the earliest business analytics implementations in Australia for one of the largest banks and the largest retailer. In 1999 she commenced research with a neonatologist from a neonatal ICU in Australia and has continued to propose new and innovative approaches for the use of information technology in neonatal intensive care and specifically the application of Big Data techniques to neonatal intensive care. She established and leads the Artemis Project, a Big Data solution for neonatal intensive care to demonstrate new data intensive solutions for conditions such as late onset neonatal sepsis, neonatal apnea and spells, retinopathy of prematurity and anemia of prematurity. Her new work also uses Big Data techniques to assess the impact of morphine on the premature neonate. She now progresses her research within the context of critical care medicine, mental health, astronaut health and military and civilian tactical training. She has been awarded over $10 million in research, consultancy and infrastructure funding. She has led the establishment of two IT start-up companies internationally and has published over 130 research publications and 7 patents internationally. She has extensive collaborative relationships with healthcare organizations, researchers and industry in several countries around the world including Canada, Australia, USA, China and Ireland. In 2013 her Artemis project was awarded the Information Technology Association of Canada (ITAC) Ingenious Award in the Not for Profit Category. In 2014 she was awarded membership in the Order of Australia, general division, for significant service to science and innovation through health care information systems. She is regularly called upon by the media as an international specialist in health informatics and Big Data. Annual Quality Congress Breakout Session, Sunday, October 4, 2015 Physiologic Monitoring in the Cloud Objective: Develop an awareness of Big Data and Cloud Computing and how these technologies can be utilized to enable new approaches for the monitoring and analysis of physiological data in the cloud.
Physiologic Monitoring in the Cloud
Carolyn McGregor AM, PhD
October 4, 2015 1
Physiologic Monitoring in the Cloud
Carolyn McGregor AM, PhDCanada Research Chair in Health Informatics, Professor,
University of Ontario Institute of [email protected]
hir.uoit.ca
@UOITHIR@CP_McGregor
Disclosure
• Nothing to disclose
DetectingInfection
Per Hour Data
3,600,000 ECG
3600 HR, SpO2
230,000 Chest Impedence
36,000,000 EEG
7000 Heart Beats
2000 Breaths
Internet of ThingsKnowledge TranslationChallenge
• Retrospective Analysis
• Not implemented in clinical practice
• Not scalable
• Either or combination of:
– Patient centric
– Condition centric
– Stream centric
Physiologic Monitoring in the Cloud
Carolyn McGregor AM, PhD
October 4, 2015 2
Artemis:Big Data Disruptive Innovation
8
Data Integration MgrKnowledge Extraction
Data Miner HIRData Mover
InfoSphere Streams Runtime
Ontology DrivenRule Modifier
Deployment Server
Alert SinkOp
QRS
BP
RR PT
FA
WT
AR
SepsisBPA
EP
WTA
HR Source Op
SpO2 Source Op
BP Source Op
CIS Source Op
PatientStream
SPAD
E ID
E
USE
R IN
TE
RFA
CE
MedicalDataHub
CIS Adapter
ConfigurationServer
CapsuleTechServer
ClinicalInformation
System
Cognos
Data Aquisition
Online Analysis
Knowledge Extraction
(Re)deployment
ResultPresentation
Stream Persistency
Artemis
McGregor, C., Catley, C., James, A., Padbury, J., “Next Generation Neonatal Health Informatics with Artemis”, Medical Informatics Europe, Oslo, Norway, pp 115‐119
Late Onset Neonatal Sepsis (LONS)
0
500
1000
1500
2000
2500
3000
3500
4000
29/06/2010 12:00:00
29/06/2010 13:00:00
29/06/2010 14:00:00
29/06/2010 15:00:00
29/06/2010 16:00:00
29/06/2010 17:00:00
29/06/2010 18:00:00
29/06/2010 19:00:00
29/06/2010 20:00:00
29/06/2010 21:00:00
29/06/2010 22:00:00
29/06/2010 23:00:00
30/06/2010 0:00:00
30/06/2010 1:00:00
30/06/2010 2:00:00
30/06/2010 10:00:00
30/06/2010 11:00:00
30/06/2010 12:00:00
30/06/2010 13:00:00
30/06/2010 14:00:00
30/06/2010 15:00:00
30/06/2010 16:00:00
30/06/2010 17:00:00
30/06/2010 18:00:00
30/06/2010 19:00:00
30/06/2010 20:00:00
30/06/2010 21:00:00
30/06/2010 22:00:00
30/06/2010 23:00:00
01/07/2010 0:00:00
01/07/2010 1:00:00
01/07/2010 2:00:00
01/07/2010 3:00:00
01/07/2010 4:00:00
01/07/2010 5:00:00
01/07/2010 6:00:00
01/07/2010 7:00:00
01/07/2010 8:00:00
01/07/2010 9:00:00
01/07/2010 10:00:00
01/07/2010 11:00:00
01/07/2010 12:00:00
01/07/2010 13:00:00
01/07/2010 14:00:00
01/07/2010 15:00:00
01/07/2010 16:00:00
01/07/2010 17:00:00
01/07/2010 18:00:00
01/07/2010 19:00:00
01/07/2010 20:00:00
01/07/2010 21:00:00
01/07/2010 22:00:00
01/07/2010 23:00:00
02/07/2010 0:00:00
02/07/2010 1:00:00
02/07/2010 2:00:00
02/07/2010 3:00:00
02/07/20104:00:00
Hourly Rows Level=0 Hourly Rows Level=1 Hourly Rows Level=2 Hourly Rows Level=3 Hourly Rows Level=4
In 1924
Cloud Computing
h //i di i /i h l d/
Artemis CloudWIHRI
McGregor, C., 2011, “A Cloud Computing Framework for Real‐time Rural and Remote Service of Critical Care”, IEEE Computer Based Medical Systems, Bristol, UK, 6 pages CDROM
Physiologic Monitoring in the Cloud
Carolyn McGregor AM, PhD
October 4, 2015 3
0
10
20
30
40
50
60
1/1/08 0:00
1/1/08 12:00
1/2/08 0:00
1/2/08 12:00
1/3/08 0:00
1/3/08 12:00
1/4/08 0:00
1/4/08 12:00
1/5/08 0:00
Date
HRV
RRV
Number of minutes low
variability/hour
LONS and Narcotics
0
10
20
30
40
50
60
1/1/08
0:00
1/3/08
0:00
1/5/08
0:00
1/7/08
0:00
1/9/08
0:00
1/11/0
8 0:00
HRV
RRV
Number of minutes low
variability/hour
DateMcGregor, C., Catley, C., James, (2012), “Variability Analysis with Analytics Applied to Physiological Data Streams from the Neonatal Intensive Care Unit”, 25th IEEE International Symposium on Computer‐Based Medical Systems (CBMS 2012), Rome, Italy
17.83
7.72
36.3838.97
45.79
53.31
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
55%
60%
7am - 7 pm Day Shift 7pm - 7 am Evening Shift
O2 saturation percent frequency using Artemis 1 Hz data collection, day 28
<85% Below 85%- 92% Target 93% - 100% Above
8.33
25.0025.00
41.67
66.67
33.33
0%5%
10%15%20%25%30%35%40%45%50%55%60%65%70%
7am - 7pm Day Shift 7pm - 7 am Evening Shift
O2 saturation percent frequency using top of the hour spot readings, day 28
<85% Below 85% - 92% Target 93% - 100% Above
OxygenAnalytics
OxygenAnalytics
NeonatalSpells
Spells Classification Anemia of Prematurity
0
10
20
30
40
50
60
HR RR
Physiologic Monitoring in the Cloud
Carolyn McGregor AM, PhD
October 4, 2015 4
PK/PD Morphine
Bressan N McGregor C Smith K Lecce L James A 2014 “Heart rate ariabilit as an indicator for morphine pharmacokinetics
Implications forClinical Engineering
Greer, R., Olivier, C., Pugh, J. E., Eklund, J.M., McGregor, C., 2014, “Remote, Real‐Time Monitoring and Analysis of Vital Signs ofNeonatal Graduate Infants”, 36th Annual International Conference of the IEEE EMBS, Chicago, USA, pp 1382‐5
Data Integration MgrKnowledge Extraction
Data Miner HIRData Mover
InfoSphere Streams Runtime
Ontology DrivenRule Modifier
Deployment Server
Alert SinkOp
QRS
BP
RR PT
FA
WT
AR
SepsisBPA
EP
WTA
HR Source Op
SpO2 Source Op
BP Source Op
CIS Source Op
PatientStream
SPAD
E ID
E
USE
R IN
TE
RFA
CE
MedicalDataHub
CIS Adapter
ConfigurationServer
ClinicalInformation
System
Data Aquisition
Online Analysis
Knowledge Extraction(Re)deployment
ResultPresentation
Stream Persistency
Artemis in Space
Stream Persistency
Knowledge Extraction(Re)deployment Stream Persistency
ResultPresentation
Data Integration MgrKnowledge Extraction
Data Miner HIRData MoverOntology Driven
Rule Modifier
Deployment Server
PatientStream
McGregor C 2013 “A Platform for Real‐time Online Health Analytics during Spaceflight” IEEE Aerospace CDROM 8 pages
Athena
Data Integration MgrKnowledge Extraction
Data Miner HIRData Mover
InfoSphere Streams Runtime
Ontology DrivenRule Modifier
Deployment Server
Alert SinkOp
QRS
BP
RR PT
FA
WT
AR
SepsisBPA
EP
WTA
HR Source Op
SpO2 Source Op
BP Source Op
CIS Source Op
PatientStream
SPAD
E ID
E
USE
R IN
TE
RFA
CE
Cognos
Data Aquisition
Online Analysis
Knowledge Extraction
(Re)deployment
ResultPresentation
Stream Persistency
Athena
McGregor C Bonnis B Stanfield B Stanfield M 2015 “A Method for Real time Stimulation and ResponseMonitoring using Big Data and
Physiologic Monitoring in the Cloud
Carolyn McGregor AM, PhD
October 4, 2015 5
Change Management andTechnology Impact Assessment
• Healthcare Innovation is not a Magic Pill– Rigorous research
– Linkages to outcomes
– Scalability/Reliability/Resiliency
– Policy Changes to Guidelines and Practices
– Impact on Healthcare Organisations
– Change Management for Integration into Care Practices
– New Information Technology Strategic Architectures
– Use of Data Governance
ArtemisClinical Applications
Questions?