CLOUD COMPUTING FOR MEDICAL RESEARCH AND HEALTHCARE Yu-Chuan (Jack) Li, M.D., Ph.D., FACMI Graduate...
-
Upload
cory-newman -
Category
Documents
-
view
215 -
download
1
Transcript of CLOUD COMPUTING FOR MEDICAL RESEARCH AND HEALTHCARE Yu-Chuan (Jack) Li, M.D., Ph.D., FACMI Graduate...
CLOUD COMPUTING FOR MEDICAL RESEARCH AND HEALTHCAREYu-Chuan (Jack) Li, M.D., Ph.D., FACMI
Graduate Institute of Biomedical Informatics
College of Medical Science and Technology
Taiwan Medical University
Taipei Medical University (TMU)
• Top private medical university in Taiwan• 6000 students, 620 faculty members, 7 colleges• Closest to the world’s highest building – Taipei 101
• Largest JCI-Accredited teaching hospitals in Taipei• 3,150 beds• Over 10,000 Out-patient visit per day
3
北醫附醫 萬芳 雙和
TMU Healthcare Group
Taipei Medical University
4
Total Faculty6,102
Students: 6,059
Alumni: 31,214
Full-time Instructor 428
Part-time Instructor 649
7 Colleges13 Departments
16 Graduate Institutes
3 TMU Hospital
s
College of Medical Science and Technology - TMU
• Department of Biomedical Informatics• 80 master and Ph.D. students
• Department of Medical Technology• 60 master and Ph.D., 300 undergraduate
students• Department of Cancer Biology and Drug
Discovery• Ph.D. only
• Department of Neuro-regenerative medicine• Ph.D. only
Medical Cloud
Care Cloud
Wellness Cloud
CitizenHealthRecord
Clinical Care
Long-term Care
Prev
entio
n
NIST Definition v.15
• Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.
Five Characteristics of Cloud
• On-demand self-service
• Broad network access
• Resource pooling
• Rapid elasticity
• Measured Service
“the kind of service that dry-lab biomedical researchers would always wanted…”
Other Terms related to Cloud• Service Model
• Cloud Software as a Service (SaaS)
• Cloud Platform as a Service (PaaS)
• Cloud Infrastructure as a Service (IaaS)
• Deployment Model
• Private cloud TMUH as an example
• Community cloud
• Public cloud
• Hybrid cloud
Private Cloud in TMUH
10 Gb10 Gb
2 Gb2 Gb2 Gb2 Gb2 Gb 2 Gb2 Gb 2 Gb
Fiber 10GbFiber 10Gb
Core Switch Engine
HIS主機 HIS主機
Core Switch Engine
Catalyst Express 500 Series
1 2
SETUP
SYSTEM
ALERTPOE
1 2 3 4 11 129 107 85 6 13 14 23 2421 2219 2017 1815 16
16X
13X
14X
23X
24X16X
1X
2X
11X
12X
Catalyst Express 500 Series
1 2
SETUP
SYSTEMALERT
POE
1 2 3 4 11 129 107 85 6 13 14 23 2421 2219 2017 1815 16
16X
13X
14X
23X
24X16X
1X
2X
11X
12X
100 Mb100 Mb
各機櫃之 Switch
100 Mb100 Mb 100 Mb100 Mb 100 Mb 100 Mb
住院Client
住院Client
門診Client
門診Client
行政Client
行政Client
醫療Client
醫療Client
連接至第一、二大樓網路主幹交換器
1
2
3
4
5
6
FANSTATUS
Power Supply 1 Power Supply 2
Catalyst 6500 SERIES
WS-SUP720-3BXL
SUPERVISOR 720 WITH INTEGRATED SWITCH FABRIC/PFC3BXL
EJECT
DISK 0
EJECT
DISK 1
CONSOLE PORT 2
PORT 1
W S - S U P 7 2 0 - 3 B X L
S U P E R V I S O R 7 2 0 W I T H I N T E G R A T E D S W I T C H F A B R I C / P F C 3 B X L
S YS TEM
S TA TU S
A C TIVE
P WR
M G M T E J E C T
D I S K 0
E J E C T
D I S K 1
C O N S O L E P O R T 2
P O R T 1
LI NK
LI NK
LI NK
R ES ET
WS-X6748-SFP
48 PORT GIGABIT ETHERNET SFP
STATUS
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
WS-X6748-SFP
48 PORT GIGABIT ETHERNET SFP
STATUS
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
1
2
3
4
5
6
FANSTATUS
Power Supply 1 Power Supply 2
Catalyst 6500 SERIES
WS-SUP720-3BXL
SUPERVISOR 720 WITH INTEGRATED SWITCH FABRIC/PFC3BXL
EJECT
DISK 0
EJECT
DISK 1
CONSOLE PORT 2
PORT 1
WS-SUP720-3BXL
SUPERVISOR 720 WITH INTEGRATED SWITCH FABRIC/PFC3BXL
S YS TEM
S TA TU S
A C TIV EP WR
M G M T EJECT
DISK 0
EJECT
DISK 1
CONSOLE PORT 2
PORT 1
LI NK
LI NK
LI NK
R ES ET
WS-X6748-SFP
48 PORT GIGABIT ETHERNET SFP
STATUS
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
WS-X6748-SFP
48 PORT GIGABIT ETHERNET SFP
STATUS
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Catalyst Express 500 Series
1 2
SETUP
SYSTEMALERT
POE
1 2 3 4 11 129 107 85 6 13 14 23 2421 2219 2017 1815 16
16X
13X
14X
23X
24X16X
1X
2X
11X
12X
Catalyst Express 500 Series
1 2
SETUP
SYSTEM
ALERT
POE
1 2 3 4 11 129 107 85 6 13 14 23 2421 2219 2017 1815 16
16X
13X
14X
23X
24X16X
1X
2X
11X
12X
Catalyst Express 500 Series
1 2
SETUP
SYSTEMALERT
POE
1 2 3 4 11 129 107 85 6 13 14 23 2421 2219 2017 1815 16
16X
13X
14X
23X
24X16X
1X
2X
11X
12X
Catalyst Express 500 Series
1 2
SETUP
SYSTEM
ALERT
POE
1 2 3 4 11 129 107 85 6 13 14 23 2421 2219 2017 1815 16
16X
13X
14X
23X
24X16X
1X
2X
11X
12X
Catalyst Express 500 Series
1 2
SETUP
SYSTEMALERT
POE
1 2 3 4 11 129 107 85 6 13 14 23 2421 2219 2017 1815 16
16X
13X
14X
23X
24X16X
1X
2X
11X
12X
Catalyst Express 500 Series
1 2
SETUP
SYSTEM
ALERT
POE
1 2 3 4 11 129 107 85 6 13 14 23 2421 2219 2017 1815 16
16X
13X
14X
23X
24X16X
1X
2X
11X
12X
Save 50-70% IT Cost Power saving 30% More efficient and effective
for resource applications
Non-Stop Intensive OLTP-type
Double-loop systemMore Security
Virtual MachinesMore Green
Business model
(Duplicability & profitability)
Operation model
(Sustainability)
Service model
(Feasibility & Value)
Me
dica
l C
loud
Care
Cloud
Wellness C
ould
Care Cloud•Service model yet to be determined
•Need more evidence Systems Clinical Trial Insurance
•Useful for the aging population•HIT development – Taiwan Experience
Night time fall - bed occupancy
Day time fall
Lifeline PNC Monitoring Centre
Carer
Bed Sensor
Night time fall - bed occupancy
Telecare, UKExample: Telecare Helping Manage Falls Risks
West Lothian
22 min fall response
4hour Scotland average
16
Telehealth Service Model
Community Care
NHII Information Platform
Institutional CareHome Care
Telehealthcare Service Center(TSC)
Emergency care center
Telehealthcare Information Platform (TIP)
LTC managementcenter
0800-008-850
members
Wellness Cloud•High possibility of Innovation
• e.g. +Social Networking• “Hey, it’s cool to stay healthy!”
• e.g. +gym, wellness, travel, sports and food industry
• e.g. + consumer electronics
Clinical Data Repository (CDR)
• Often a backend of
Electronic Health
Record (EHR) for
efficient patient-
level data access
• High potential for
clinical research
NHIRDB• NHIRDB (National Health Insurance
Research Database)
• 12 years of de-identified claim database for 23 million people
• Cohort DB (Five 1-million people groups for 13 years)
• Disease-specific DB (16 disease groups)
• Random sample DB (outpatient 1/500, inpatient 1/20)
• generates >100 research papers a year
What is MEUC (pronounced as “Milk”)
• MEUC (MEDICAL END-USER COMPUTING), an system that provide end-users a simple to use interface to access “Big Data” for biomedical research in a Cloud Computing framework.
Nutrition to researchers(Proudly made in TMU)
The Size of MEUC
Year OPD IPD ER Drug Exam Lab Procedure
2007 2,810,380 84,415 164,313 8,377,165 878,392 12,185,057 963,163
2008 3,187,302 102,480 185,755 9,635,517 1,150,302 13,445,714 1,032,721
2009 3,666,632 164,355 252,414 11,370,756 1,368,141 11,435,334 1,274,819
2010 3,058,730 73,759 184,890 10,149,577 694,654 9,440,527 1,370,797
2011 3,580,217 80,253 188,372 10,607,887 769,530 7,929,639 1,442,294
Total 16,303,261 505,262 975,744 50,140,902 4,861,019 54,436,271 6,083,794
16M 505K 975K 50M 4.8M 54M 6M
DDQ-Disease and Disease Q value• From NHI Database 2000 、 2001 and 2002 Data.
• Visits Count more than 230(total population 1/100,000) Disease , It means exclude rare diseases , we can get 4,000 disease.
• 16,000,000 Q value of each Group of 20 Group.
CTM
Hospitals
URS(non-radiology images, signals
and report)PIS
(Pathology Information System)
EMR (Electronic
Medical Record)
PACS(Radiology
images and report)
LIS(Lab Information System)
MEUC (Medical End-User
Computing)
CBIS(Bio-repositoryInformation
System)
Tier 2 & 3Bio-repository
Bioinformatics Platform
Drug-DiscoveryPlatform
IAS(IntelligentAcquisitio
nSystem)
Translational Bio-Informatics Road Map
Tier 1 & 0Bio-repository
EMR Integration Gateway
CPOE(Computer Physician Order Entry)
PCS(Proactive Consent System)
from Prof. Yu-Chuan (Jack) Li, 2009-12, last revised 2012-07-24
Translational Bio-Informatics Road Map
from Prof. Yu-Chuan (Jack) Li, 2009-12, last revised 2012-07-24
Translational Research Cloud
Researchers
8 Center of Excellencefor Cancer Research
National Cancer InstituteNCI
Lab & Exam DB
Exome, SNP, Microarray, Proteomic 2D page…etc.
Radiology, Pathology(MTA...)
Nuclear Medicine Image
Therapeutic Information DB
Data Mining Tool Box
Systems Epidemiology for Cancer Research
CaBIG Gateway
Secondary Cancer Research Database
Master Patient Index (Pseudo ID)
PrivatePublicLimited
Electronic Medical Record
NHIRDB Cancer Data
Tissue Bank Data
Cancer Registry Data
Cancer Screening Array Data
Bio-signal DB
Privacy & Security Firewall
UI, Access
Control
Challenges• ELSI and privacy issues
• IRB or no IRB or a Joint IRB• More complex with more partners
• Lack of data standard for EMR data• caBIG, caDSR (Cancer Data Standards Registry and
Repository) for hosting and managing metadata)
• Scale and complexity of EHR• High demand of computational resource to maintain
multiparty private computation (shared results without sharing data)
• Not RCT clinical trials (but much more available)
Cloud Computing for Major Hospitals• Virtualization (IaaS)
• Cut the maintenance cost in half
• Much less server room space and much greener
• More flexibility and portability of services
• Service-Oriented Architecture (SaaS)
• An overhaul of the basic system architecture
• Highly flexible and efficient new architecture
• Fast deployment of new applications
• Web-friendly
Cloud Computing for Personalized Health Care
• PHR or ePHR (electronic Personal Health Record) is at the the core of the Next-Generation Health IT
• PHR =• Personal part of the EMR from all the providers • + self-measured bio-signals • + self-entered health related information like family history,
exercise, food consumption, food allergies, OTC drugs, cigarette consumption…etc.
• PHR will be the basis of Personalized Healthcare
Personalized Medicine
• It is estimated in 2014, a personal
Genome can be sequenced under
$1,000 USD
• 3 billion DNA and 33K genes
more than 100K proteins
metabolic pathways
all the functions of body
Some estimated 4GB to store all the
short-reads (before compression)
Personalized Medicine (cont.)
• Need a place to • store it• review it• make sense out of it by linking to a person’s health information
• PHR on the Cloud will be the ideal place
Cloud Computing for Rural Heath Care
• The lack of IT resource of rural health stations and small hospitals in China (>16,000)
• Pioneered since 2007 by Steve Chan (designer of Cray-2)
• Deployed into two medium-sized cities in the western part of China (200 health stations, 800 care-givers)
• A new sustainability model for HIT in developing countries and resource-poor areas
Conclusion• Cloud Computing will change the face of Biomedical
Research Data Service
• Cloud Computing, with privacy-enhanced, could change the future of HIT delivery in developing countries and resource-poor areas
• Fits the needs of many healthcare sectors due to flexibility and cost-effectiveness
• Cloud Computing will be a “liberator” for scalability/accessibility limitations