the vgr way to handle big that will get bigger · • big data extends beyond data that you control...
TRANSCRIPT
The VGR way – To handle Big that
will get Bigger
Mikael Wintell., Chairman DICOM WG 26 (Digital Pathology),
Member IHE, HL7, ISO TC215, eHealth department,
HSA, Region Västra Götaland, Sweden
Lars Lindsköld Ph.D., eHealth department, HSA, Region
Västra Götaland, Sweden
2
29 in house X-RAY departments
4 independent imaging services
One Teleradiology solution outside VGR
( Barcelona / Sydney)
One million exams per year / Including SR/ XDS-I /SC
241 modalities connected Radiology
14 PACS DB
14 RIS DB
Dental Care 1500 labs
3 Clinical Physiology departments
2 Screening projects ( AAA an SCAPIS)
Stored 910 TB Data since 2006
More than 970 millions independent DICOM objects
“Golden Record”- VGR
WHY?
Big data that will get bigger
within the Healthcare society
?
PREPARE TO SHARE
IT´S ALL ABOUT
CAREI’M SHARING WHO
CARE´S SHARING IS CARING
• To make the most of Big Data we need to ensure that the
source information is trustworthy
• But the complexity and sheer volume of data means that
traditional methods of governance is no longer feasible.
• We need to think out of the BOX!
Big Data
• When we added transaction details,
multichannel interactions
• Syndicated data from sources such as
”IRS”, NPS, BFR, SCAPIS, BOIC etc
• When and if we add data from social
media
When the ”Golden Record”
becomes Big Data in Healthcare
• Big data extends beyond data that you control in your applications
and data warehouses.
• Potential insights and conclusions can be buried in the ”mess” of
structured, unstructured documents, spreadsheets, reports, email etc
• What we know for sure, the volume of data concerning patients and
their equipment continues to grow each second (24/7/365).
• Another dimension, patient data is often captured in high-velocity
streams. If we can’t do real-time analysis important insights may be
lost and also ”spot on decisions” (neonatal ICU sensor data)
More data = More *understanding?
Not necessarily!
Problem:
Possible resolution Microscope: 0,2 μm
Uterus: approx.200 g tissue ≈ 200 cm³
= 250 000 000 000 000 000 voxels
At 24 bit color: approx.750 petabyte information
Approx.1000 samples/year-> 750 exabyte ( VGR) “As of May 2009, the size of the World's total Digital content has been roughly estimated to be 500 billion gigabytes, or 500
exabytes.”
- What shall we choose?
Evidenced based care
Personalized Care
Big Data in Medicine
The way we in the future will
handle and organize our
digital data in Healthcare –
might have the biggest
impact on human welfare
ever seen
To know as much as possible
on many ( Evidence care )
Could we better serve you on
an individual base
(Personalized care)
Production on
healthcare data
VERSUS
The need for
information in
Research
N = ALL
STORAGE DISTRIBUTION
Box Structur
Nationell Security
Infrastructur
SR basic, SR
advanced
SR to SC
Structured reports
to Secondary
Capture
REGIONAL
ARCHIVE
LOGICAL
OWNERSHIP
XDS
BROKER
MRP = Multi RIS
Production
Image
Production
API
External provider
and user
Split Information
Producer/
Consumers
Example:
National
Patient Summary
Example:
SECTRA RIS/
PACS
DICOM / HL7 / IHE
SPIDER
WebView Centricity
Image Portal ?
Others ?
DICOM
XML ( IB )
New Topic
ILM
Image Life
Management
RIS
Interoperabilitet
Way In / Out
VIRITUALIZATION
System
Data
Information
Profession
People
FoUU
New Specialities
Without DICOM /
HL7
Information Model
Framworks = Generic,
Radiology other
Additional tagging
Tag Morphing
DATA TIERS
Quality Register´s
HCP
The tension between:
Patients Integrity Patient Safety
The tension between:
Informatics Federations two directions
“Faster, better, cheaper” “Fear, uncertainty and doubt,”
