Transcript PACS & Pathology

Evolution of PACS as a
Model to Predict Adoption of
Digital Pathology
Joe Biegel
VP, Product Management
McKesson Medical Imaging Group
October 23, 2007
A Brief History of PACS
US Army teleradiology
project starts.
1983
Siemens/Philips start
developing PACS concepts.
EuroPACS starts annual
meetings.
US – Baltimore Veterans
Hospital PACS.
UK – Hammersmith and
Conquest PACS projects.
1984
Original concepts: Dr. T. Iinuma,
“Image processing in clinical
medicine – considerations of a
system”, Bull., Tokyo Women
Med. 44, 152 (in Japanese).
US National Cancer
Institute funds UCLA
group under Professor
Bernie Huang.
1974
1985
Baumann and Gell follow up survey –
177 large PACS identified worldwide.
2000
1992
Multiple regional and
national projects
initiated. Worldwide
mass adoption
begins in earnest
2002
1970’s
1980’s
1990’s
2000’s
Roger Baumann and Gunther Gell identify 23 large
PACS worldwide with over 100 PACS projects.
Dr. Samuel J. Dwyer III describes PACS
ideas at a Seminar at the University of Iowa.
Dr. Judith M. Prewitt introduces the term
PACS. Cardiovascular radiologist Dr. Andre
Duerinckx describes the far reaching future
benefits of PACS for a hospital.
Berlin BERKOM
project started.
NATO conference in France includes topic on Imaging Networks.
SCAR celebrates 10th annual meeting.
1981
1986
1990
1995
1st International Conference
on the PACS, Newport
Beach, California.
PRIMIS created at VUB
Brussels – to study
applications of digital imaging.
1982
1987
PACS Market Evolution
1985
1990
2000
Replacement
Mass Adoption
Market
Early Adopters
Research Pioneers
Single Modality Applications
Solutions
Offered
Departmental PACS
Multi-Modality Viewing Apps
Departmental RIS/PACS
DICOM + HL7
ACR-NEMA 2.0
Standards
DICOM 3.0
ACR-NEMA
Mainframe Computers
Technology
Enablers
Enterprise RIS/PACS
9-track tapes
IHE
TCP/IP adoption
Optical Storage
Jukeboxes
Affordable UNIX
workstations
Optical Storage Media
Hi-Res Displays
HSM Systems
Voice
Recognition
NAS / SAN
Storage H/W $ 
The Evolution of Imaging IT Market Segments
1985
1990
2000
2010
Replacement
Optical
Imaging
Mass Adoption
Research Pioneers
Cardiology
PACS
Market
Early
Adopters
Replacement
Mass Adoption
Research Pioneers
Early
Adopters
Replacement
Radiology
PACS
Market
Mass Adoption
Research Pioneers
Early
Adopters
Early Perceived Value
Historical Radiologist Workflow
Films with
masters are
delivered to
Radiologist
Clerks/Tech match
up films to be read
with patient’s
master jacket
Rad hangs
films on
view box
Rad enters
MRN into RIS,
finds patient,
reviews results
Rad hangs
previous
for review
Is there
previous in
master
1
No
Rad
bar codes
Rad bar codes
requisition
into
requisition into
dictation system
dictation
system
1
Rad dictates patient name,
order, and accession
number in dictation system,
and dictates report
Rad places
films back into
daily folder and
signs bag
Rad places signed
daily with patient
master jacket aside for
pickup by clerical staff.
Radiologist Workflow with RIS/PACS
Rad opens unread
worklist on PACS in
dictate mode and opens
first exam to be read
1
Relevant prior
studies automatically
display
Patient and order level
information sent from PACS
to RIS with no patient
search required
Dictation automatically
initiated. Order information
from the RIS automatically
sent to dictation system.
Rad dictates report,
and clicks next study
Rad clicks on clinical
history icon to review
patient’s reports and
results from HIS/RIS
Next study is loaded.
Dictation is initiated, and
previous exam is
marked dictated in PACS
1
Pathology progress,
as compared to Radiology
 The evolution of Digital Imaging in Radiology
offers significant lessons for Pathology
 Parallels with a later start,
… but likely faster progress
─
Radiology in 1995 vs. Path in 2005
─
Radiology today vs. Path in 2010
 Moore’s law helps
 No need to repeat the same pitfalls
Source: Bruce Beckwith, MD – 2005 DICOM International Conference – Budapest, Hungary
Radiology c. 1995 -vs- Pathology c. 2005
Radiology
Pathology
Analog image capture
Analog image capture or no image capture
Single modality to multi –modality focus
Single modality
DICOM standards- early formation of IHE
Emerging standards
Separate workflows – RIS, Reporting,
Separate workflows – APLIS, Reporting
Radiology focused
Images hard to interpret for non
pathologists
Storage needs seem overwhelming
Extreme storage needs
Digital Imaging is more expensive than
analog
Digital Imaging costs money
“Island” workflow
Unstructured, collaborations, distribution
Most of spending is still on film
Modest budgets
Radiology Now -vs- Pathology in 2010
Radiology
Pathology
Digital Acquisition is standard
Digital acquisition growing
Automatic meta-data capture
Streamlined image capture with metadata
CAD assists Radiologist’s interpretation
CAD interpretable portions
Many patient requests
Large transactions per study
Large and constantly growing storage
needs
Large storage needs, but cost barrier
decreasing
Digital imaging saves money, and is
required to make money
Digital pathology adds value and is
required to make money
Large budgets
Budgets expand into IT space
Radiology outside the dept. of Radiology
walls
Remote pathology, protocols, strong links
to clinical practice.
Driving adoption is key
 Automating the current workflow is only the first step
─
Both physical slide based workflow and digitized slide workflow will
co-exist for a long time
─
Metadata (ie. patient demographic info, block/slide info) needs to
be assigned to digitized WSI images
─
Manual data entry must be minimized whenever possible if WSI
technology is to be incorporated into and ultimately replace
existing physical slide based workflow
 Economic drivers of digital pathology are evolving
─
Factors from outside the department will dominate
─
Standard of care, costs & reimbursement are the keys
 Solution requirements will also evolve
Pathology Workflow
Present & Future
Cassettes/Blocks
Histology
Laboratory
Gross
Specimen
10110101010101010
Pathologist Orders: Recuts,
Levels, IHC/Special Stains,
Ancillary Studies
WSI unique ID # for each WSI slide
1010101010101010101
101010101
10110010101010101010101010101
Patient Block Slide
Data attached to
WSI Unique ID#
Region of Interest
(ROI) jpeg images
(Direct Operation
3rd Party Imaging
Apparatus)
10101010101010101010101010101010
3rd Party Whole Slide
Imaging (WSI) Apparatus
Analog
Optical
Image
10101101
Physical Slide
0101010101010101010010101010101010101010101101010101010101010101010101010101
Digitized Whole Slide Images (WSI)
attached to WSI Unique ID#
Source: Sourav Ray MD, et al: CAP, 2006
101010110100101010101010101010
Region of Interest
(ROI) jpeg Images
APLIS Remote
Operation 3rd
Party Imaging
Apparatus
(TWAIN)
10101010101010101010
Load with WSI Unique
Slide ID#
PACS
10110010101010101
101010101010101010
Patient/Case/
Block Data,
Prior Reports,
Images (ROI)
101010111011
and WSI)
WSI Images as
10110010101010101010101
requested per case
101010101010101010
Order
s
101010101010110
10101010101
10101010101
Request WSI Images
Via WSI unique ID#
Request ROI Images
101010110100101010101010101010101010
Physical
Slides
01010101010
Slide
Storage
101010101
Block Data, Gross
Description,
Gross Images
H&E Data, IHC Data,
Flags, Comments
WSI or not WSI?
Annotated
Whole Slide
Images (WSI)
Metadata
Annotation
Engine
Future Pathology Workflow
 The vision of automating the current workflow is
a noble goal, but it is only the first step
─
The true value of Digital Pathology will have to be
defined in terms of faster, improved Patient Care
 Going Digital will change the practice of
Pathology and will lead to a new, yet undefined
workflows
─
Standards, IHE will help: necessary but insufficient to
drive wide-scale adoption
DICOM in Pathology
 Visible Light Supplement 15
─
Included support for gross images, microscopic
images, x,y.z position, lab accession numbers
 Working Group 26
─
Evolving to include whole slide imaging
 Growing number of imaging products targeting
pathology are DICOM compliant
 US Veteran’s Administration:
─
Pathology imaging vendors must be DICOM
compliant
IHE-J in Pathology –
Workflow/Data Structure
Study
Series
Image
Macro
Image
Order
Sample
Block
Sample
Slide
Slide
Block
Sample
Source: Tofukuji * Okuno – 2005 IHE –Japan Pathology WG
Slide
Diagnostic Imaging Cockpit
= ‘vision’ under pressure
Departmental PACS diagnostic tools
Collateral content
3D, Fusion, CAD
Orthopaedic extensions
Diagnostic Imaging Cockpit
Enterprise Worklist
Voice Recognition
Departmental management &
Structured Reporting: eg OB,
reporting tools
ER discrepancy & critical results
EMR Integration
reporting
The Pathology “Cockpit” – current focus
Worklist &
Case History
Gross / Low Res.
Image Analysis
Digital Slides
A
B
A
This is an example text bocks for the gross image shown
above. This text is so small that it is unlikey that anyone
can read it. Can you? I really doubt it.
This is an example text bocks for the gross image shown
above. This text is so small that it is unlikey that anyone
can read it. Can you? I really doubt it.
Source: Aperio Technologies
B
Report
Pathology  Next stage
Advanced diagnostic tools
3D, Fusion, CAD
Muti-spectral imaging
Test sub-routines (histology, etc.)
Pathology Cockpit
Enterprise Worklist
Voice Recognition
Departmental management &
Structured Reporting
reporting tools
Critical findings / ER ??
EMR Integration
Departmental Productivity Reporting
Synergies with Diagnostic Imaging
 Leverage image management
infrastructure
─
Cost and expertise synergies will
speed adoption
 Routine access will likely
create clinical use cases which
will drive the adoption
─
The extent to which these use
cases impact outcomes will
impact market acceleration
 Access needs vary
─ Pathologists want access to
diagnostic image data
─ Radiologists have interest in
access to summary data and no
interest in WSI access
Value Points = Business Control Points
 Modality = Data Acquisition
 Storage = Physical Storage Devices (H/W)
 Image Access Points = Clinical Application(s)
 Workflow = Order  $
 Data = Data-mining
 CAD – address the volume – may be more
formative than in Radiology
 Financial benefit needs to become more clear
Organic evolution is a slow, painful
process with many side-tracks
 Integration and interoperability with
other systems is key to adoption
─
Isolated systems, IT-silos are short-lived,
with limited clinical benefits
 Don’t repeat the evolutionary
sidetracks from the history of other
clinical disciplines…
 Use PACS as a starting point
and build on it !!
Summary
 Pathologists need PACS now
─
─
value is immediate
Pathologists should have PACS access now – the patient record
needs to contain relevant imaging data
 If you want to accelerate the adoption-curve of digital
pathology:
─
‘Hook’ your digital Pathology plans to your PACS
 It is likely that early workflows will leverage other
specialties, BUT these will inevitably be transient
solutions with the “real” Digital Pathology workflow to
follow
─
Pathology dept specific workflow solutions are needed and must
have digital path imaging as part of the workflow now.
Questions?