Transcript R-Horn_What`s New in DICOM 2008

What’s New in DICOM
Robert Horn, Agfa Healthcare
SPIE Medical Imaging, 2008
DICOM Administrative Information
• DICOM is also ISO 12052.
• The most recent published version, Supplements,
Change Items, etc. remain at NEMA
• MITA (Medical Imaging and Technology Alliance)
• is the new name for the medical division of NEMA.
• 2008 Version is published
• ftp://medical.nema.org/dicom/medical/2008
2
SPIE February 2008, San Diego
Multiframe Image SOP Classes
• Multiframe objects to support 1000+ slice studies
• Image header supports functional group attributes
changing during acquisition
• Dimensions allow multiple views of data
• Common structure used for all new image IODs
• Sup 110 Ophthalmic Coherence Tomography (Adopted
August 2007)
• Sup 116 3-D X-ray (Adopted January 2007)
• Sup 117 Enhanced PET (in Letter Ballot)
• Sup 43 Enhanced Ultrasound (in Public Comment)
• Sup 125 Breast Tomosynthesis (in Public Comment)
3
SPIE February 2008, San Diego
Multiframe object structure
N Objects, N Headers
N Frames, One Header
Fixed Header
4
Per-frame header
Dimension data
SPIE February 2008, San Diego
Pixel data
The Per-Frame Header
Other attributes
Shared Functional Groups Sequence
> Functional Group A
…..
Functional Groups
with attributes constant
for all frames
> Functional Group K
Per-frame Functional Groups Sequence
Item 1 (Frame 1)
> Functional Group B
> Functional Group C
Sequence of
Functional Groups
for each individual frame
…..
> Functional Group M
Item 2 (Frame 2)
…..
> Functional Group B
> Functional Group C
…..
> Functional Group M
Pixel Data
…..
Frame 1
Item n (Frame n)
> Functional Group B
> Functional Group C
…..
> Functional Group M
Other attributes
5
SPIE February 2008, San Diego
Frame 2
…..
Frame n
Sup 119 Frame Level Retrieve
• If we have 1000+ slice image objects, how do we access
the slice identified in the report as the Key Image?
• Or, how to access a short video clip of a long endoscopy video
image object?
• Many similar use cases needing one frame or a small subset of
frames from large multiframe image
• Purpose of retrieval may not be diagnostic; compression OK
• Proposed new Retrieve SOP Classes for subsets of
frames
• Request can include a preferred transfer syntax (i.e., type of
compression)
• Returned object is a derived image with the requested subset of
frames, but with a different Instance UID
In Letter Ballot
6
SPIE February 2008, San Diego
Sup 43 Ultrasound
• Pixel frames may
represent echo power,
doppler shift, or
doppler variance
• Multiple frames (up to
3) can be blended for
proper display
• An enhanced blending
pipeline is required to
render images
Real World Value
Mapping
Data Frame
Modality
LUT
(identity)
Real World
Value +Units
VOI LUT
(optional)
Presentation
LUT
(identity)
Gray Scale
Output as
P-Values
Convert to
RGB (R=B=G)
Primary
RGBA
Palette
Color LUT
RGB (Input 1)
RGB (Input 1)
(optional)
Alpha
OR
OR
Alpha, Beta
(Constants)
MSB’s
Blending
Operation
Opacity
LUT
(optional)
LSB’s
ICC Input Profile
Alpha 1
Alpha,
Beta
LUT
outputs
OR
Alpha 2
Multiply inputs and
add the result
(Alpha 1) x Input 1
+
(Alpha 2) x Input 2
RGB
C
L
A
M
P
RGB
Profile
Connection
Space
Transformation
Color Output
as
PCS-Values
(1-Alpha), Alpha
Alpha
Data Frame
Data Frame
Modality
LUT
(identity)
Modality
LUT
(identity)
VOI LUT
(optional)
VOI LUT
(optional)
Secondary
RGBA
Palette
Color LUT
RGB (Input 2)
(optional)
Real World Value
Mapping
Real World
Value +Units
Real World Value
Mapping
In Public Comment
7
RGB
SPIE February 2008, San Diego
Sup 111 Segmentation Image
• Derived image object
• Multiframe, using functional group mechanism
• Multiple segments per object
• Each segment linked to a categorization
• Pixels show presence of category at pixel location
• Binary (1-bit/pixel) or fractional (probability or occupancy)
• Segments can be 2-D or 3-D
• 3-D segmentation is typically in same Frame of Reference
as source image
• Segments can be displayed as overlays on source image
Adopted August 2006
8
SPIE February 2008, San Diego
Sup 132 Surface Segmentation
• Segmentation using 3-D surface mesh (rather
than voxel volume)
• First use of mesh technology in DICOM
• Will also be used for future volumetric work, including
implant specification and surgical planning
In Public Comment
9
SPIE February 2008, San Diego
Sup 112 Deformable Spatial Registration
• Composite Object that supports multi-image
alignment
• 3 step algorithm
• Pre-deformation matrix transform
• Deformation using 3D grid of 3D deformation vectors
• Post-deformation matrix transform
• Facilitates further image processing, especially
multi-modality image fusion and image pasting
applications
Adopted August 2006
10
SPIE February 2008, San Diego
Structured Reporting
• Additional Templates for standard SR SOP Classes
• Sup 128 Cardiac Stress Testing (Frozen Draft – passed Ballot,
awaiting final SNOMED and IEEE codes)
• Sup 129 Cardiac Electrophysiology (in Public Comment)
• Templates for specialty SR SOP Classes
• Sup 127 CT Radiation Dose (Adopted October 2007)
• Sup 130 Ophthalmic Refractive SR (Adopted January 2008)
• Sup 126 Colon CAD (Public Comment)
• Harmonization with HL7
• Sup 114 DICOM Encapsulation of CDA Documents (Adopted
January 2007)
• Sup 135 SR HL7 CDA Mapping (Public Comment)
• Mapping is only for Basic Imaging Reports
11
SPIE February 2008, San Diego
Sup 107 Substance Administration
• Patient safety issue – “five rights” checking for contrast agents
• Point of use verification for substances often “out of the loop” of hospital
pharmacy systems
• Three new DICOM services
• Typically Modality as SCU interacting with HIS via gateway SCP
• Query: Look up product characteristics
• Given a barcode ID (from contrast agent bottle), get name, active
ingredient, concentration, etc.
• Query: Get substance administration approval
• Based on agent, patient, route of administration; get approval or warnings
• Report contrast used to HIS Medication Administration Record
Adopted January 2007
12
SPIE February 2008, San Diego
Sup 96 Unified Procedure Steps
• Simplification and merger of Worklist Management and
Performed Procedure Steps
•
•
•
•
•
•
1:1 relationship between scheduled and performed steps
A single IOD that combines attributes of both
C-FIND to query worklists
N-CREATE to create new scheduled procedure steps
N-GET/N-SET to update
N-EVENT-REPORT to monitor and report progress
• Intended to meet the needs of
• Radiation Therapy Scheduling and Monitoring
• Scheduling and Monitoring CAD systems
• Proposed new scheduling activities
Frozen Draft for Trial Implementation
13
SPIE February 2008, San Diego
Sup 74 Radiotherapy Worklist
• Supports automated worklist management and machine
parameter verification for the integration of multi-system
radiation therapy planning and treatment
• First usage of Unified Procedure Step
• Also defines three new objects / SOP Classes
• RT Beams Delivery Instruction (Composite)
• RT Conventional Machine Verification (Normalized)
• RT Ion Machine Verification (Normalized)
• Real implementation and evaluation needed – in
conjunction with IHE Radiation Oncology
Frozen Draft for Trial Implementation
14
SPIE February 2008, San Diego
Sup 122 Specimen Identification
• Support for pathology lab workflow, specimen-based
imaging
• Gross specimens, blocks, vials, slides
• Image-guided biopsy samples
• Retires current Specimen Identification Module, adds new
Specimen Module at image level of hierarchy
• Update to Modality Worklist to allow Specimen Module
• Enables automated slide scanning devices to fully populate
header
• Update to Modality Performed Procedure Step to identify
imaged specimen
• Allows LIS to track images for specimens
In Letter Ballot
15
SPIE February 2008, San Diego
Specialties
• Biomed engineering
• Supp 124 Communication of Display Parameters (in
preparation)
• Administrative support for monitor calibration
• Surgery
• Sup 131 Implant Description (in preparation)
• Supports 3D model of bone-mountable implants
• Dentistry / Radiology
• Sup 123 Structured Display (in preparation)
• Captured screen layout of multiple images (with
presentation states) for dental series, radiology
conference, etc.
16
SPIE February 2008, San Diego
Sup 118 Application Hosting
• Experimental trials are underway
• February 2008 meeting to prepare public
comment text for March 2006 WG-06 meeting
• New approaches to DICOM Information Object
exchange
• Native (shared memory) using DICOM encoding
• Native Model
• DOM-like model defined using DICOM terminology
• Abstract Model
• DOM-like model of generic image attributes
• Usable for any image type (TIFF, DICOM, etc.)
17
SPIE February 2008, San Diego