Fundamentals of CR, DR and PACS
Transcript of Fundamentals of CR, DR and PACS
1
Fundamentals of
CR, DR and PACS
J. Anthony Seibert, Ph.D.
Professor of Radiology
Sacramento, California
Learning Objectives
• Explain technology of available and future digital radiography technology
• Understand underlying system operation and characteristics
• Discuss Exposure Indices and the new international standard for CR and DR
• Illustrate integration with PACS architecture and DICOM
2
DIGITAL IMAGING IN RADIOLOGY
• Digital imaging is an essential component of Electronic Imaging, Telemedicine and Remote Diagnosis
• Steps for digital imaging – Acquisition
– Display
– Diagnosis
– Distribution
– Archive
The “Big” Picture
• “Digital” Radiology
MRI
CT Nuclear Medicine Ultrasound
Interventional Angio
PACS Digital Radiography
3
Imaging Exams
• Projection imaging 70%
• Fluoroscopy 3%
• Computed tomography 8%
• MRI 6%
• Ultrasound 10%
• Nuclear Medicine 3%
Medical Imaging Modalities
• Common thread
– Digital data (and lots of it!!)
• Problems
– Proprietary structures
– Unknown data format
• Solutions
– DICOM and PACS
– HL-7 and RIS
– Networking and Informatics
– IHE integration “profiles”
4
But First…… CR & DR
• CR: Computed Radiography using photo-
stimulable phosphors and passive detection
• DR: Direct / Digital Radiography using a
variety technologies and active detection
• An intrinsic part of the PACS
• Historically, the last electronically integrated
• Many advances in the past decade
Digital x-ray detector
Transmitted x-rays
through patient
Charge collection
device
X-ray converter
x-rays electrons
Analog to Digital
Conversion
1. Acquisition
2. Display
3. Archiving
Digital
processing
Digital to Analog
Conversion Digital Pixel
Matrix
5
Exposure Latitude
Log relative exposure
Sig
na
l o
utp
ut
10000:1
Digital
Analog versus Digital
Film
100:1
Spatial Resolution
Digital sampling
MTF of pixel aperture (DEL)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7 8 9 10 11
Frequency (lp/mm)
Mo
du
lati
on
500 mm
200 mm
100 mm
Detector
Element,
“DEL”
Sampling
Pitch
Fourier transform of Rect (Dx) = sinc(Dx)
Cutoff frequency fC = (Dx)-1
With sampling pitch = sampling aperture, Dx
Nyquist Frequency fN = (2 Dx)-1
6
Storage Phosphors
Slot-scanner
Optical Lens
CsI
Gadox
CsI
CCD
Flat-panel
a-Si
Indirect
Direct Selenium
Gadox
Cesium Iodide
Fuji, Agfa, Kodak, Konica….
Hologic, Toshiba, Shimadzu…
Canon, Carestream
GE, Trixell, (Philips, Siemens..)
Varian, others….
Imaging Dynamics, Imix,
Swissray, Wuestec
Delft Dx Imaging, LoDox
CMOS
x-Si Indirect CsI Bioptics
Photon counters
Gas detector
Solid State Si
Direct
Sectra
X-Counter
Digital Radiography 2012: detectors and manufacturers
BaFBr
CsBr
Digital Radiography common themes
• Separation of Acquisition, Display, and Archive
• Wide dynamic range
– ~0.01 to 100 mR (~0.1 to 1000 mGy) incident exposure
• Variable detector exposure operation
– 20 to 2000 “speed class”
• Appropriate SNR & image processing are crucial
for image optimization
7
Available digital radiography technology,
2012
• CR: Photostimulable Storage Phosphor (PSP) – Cassette-based detectors/readers
– Flying spot mechanical changers
– Line scan integrated detectors
• CCD: Charge-Coupled Device – 2-D lens coupled systems
– 1-D slot-scan systems
• Thin-Film-Transistor (TFT) flat panel – Indirect detection (scintillator)
– Direct detection (semi-conductor)
– Portable wireless implementations
Future “in-progress” technologies
• Hybrid direct / indirect flat panel TFT with variable gain
• Complementary Metal Oxide Semiconductor (CMOS)
• X-ray “light valve” Liquid Crystal and reflective scanner detector system
• X-ray photon counters based on gas or silicon strip detectors
8
PSP Radiography (CR)
• Currently the major technology available for large field-of-view digital imaging
• Based upon the principles of photostimulated luminescence; 20+ years of experience
• Operation emulates the screen-film paradigm in use and handling.. (flexible but labor intensive)
• Manufacturing trends: – Smaller, faster, less expensive
Computed Radiography “reader”
Information panel
Plate
stacker
Fuji
Agfa
Konica
Carestream
Various capabilities, sizes, throughput
9
Phosphor Plate Cycle
PSP
Base support
reuse
plate erasure:
remove residual signal
light erasure
plate exposure:
create latent image
x-ray exposure
plate readout:
extract latent image
laser beam scan
“Direct” Radiography (DR)
....refers to the acquisition and capture of the x-ray image without user intervention (automatic electronic processing and display)
– “Indirect” detector: a conversion of x-rays into light by a scintillator, and light into electrons for signal capture
– “Direct” detector: a conversion of x-rays to electron-hole pairs with direct signal capture
10
Flat panel portability
• Initial products introduced by Canon
– Tethered, thick profile
• Wireless products now on the market
– Trixell, Carestream, Canon, Source one…
CR and DR mobile radiography
Tethered cassette CR reader / processor
11
Wireless DR cassette
• Integrated
• Battery powered
• On-board computer
and processing
Point of service direct imaging
• 14x17inch
cassette…
12
• Preview
image in
2-3 s
Point of service direct imaging
• For Processing image…. 15 seconds
• QC
• Annotation
• Send
Point of service direct imaging
13
DR replacement trends
• Passive for active detector technology
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.0 0.5 1.5 2.5 3.5 4.5 1.0 2.0 3.0 4.0 5.0
Frequency (lp/mm)
Modula
tion
Spatial Resolution – MTF
a-Selenium: 0.13 mm
Screen-film CsI-TFT: 0.20 mm
CR: 0.10 mm
CR: 0.05 mm
14
Detective Quantum Efficiency (DQE)
• A measure of the information transfer efficiency of a
detector system
• Dependent on:
– Absorption & conversion efficiency
– Spatial resolution (MTF)
– Conversion noise & electronic noise
– Detector non-uniformities / pattern noise
– Not necessarily indicative of clinical performance
2 2
out
2
in N
SNR MTF( )DQE( ) =
SNR NPS ( )
ff
f q
DQ
E(
f )
Spatial Frequency (cycles/mm)
0.0 0.5 1.0 1.5 2.0 2.5 0.0
0.2
0.4
0.6
0.8
Detective Quantum Efficiency Radiography
CsI - TFT
Screen-film
CR Conventional
a-Se - TFT
CR “dual-side”
15
High DQE does not “guarantee” good
image quality
• Appropriate radiographic technique is required
• Optimization of acquisition technique – kV, mAs, SID, filtration, anti-scatter grid
• For similar acquisition techniques and grid use, the SNR requires radiation dose proportional to DQE-1
• “Effective DQE” concept takes into account clinical situations (magnification, grid)
Attribute CR DR CCD
Positioning flexibility **** ** **
Replacement for screen/film **** ** **
DQE / dose efficiency ** *** **
Patient throughput * *** **
X-ray system integration ** **** ****
Access to advanced
technology applications ** **** ***
Cost for comparable image
throughput *** ** ***
Radiographer ease of use
(manufacturer dependent) * *** **
16
31
• CR & DR systems have variable speed, wide dynamic range, and internal signal scaling
• Consistent (and often inconsistent) image appearance eliminates exposure feedback loop
• There is no direct link between image appearance and detector “speed class”
• Overexposures can easily be unnoticed, resulting in needless overexposure to the patient
• Underexposures have increased image noise that can reduce diagnostic accuracy
Standardized Exposure Index for Digital Radiography – Technical Issues
32
Screen-Film system indicators
Traditional screen-film systems use overall film density as an exposure indicator
Direct feedback to the technologist regarding exposure
J. Shepard and M. Flynn
17
33
CR & DR system indicators
CR & DR systems use image processing to align the grayscale with the signals
Direct visual cues (dark/light) are lost regarding exposure
J. Shepard and M. Flynn
34
Noise
The image processing adjusts the grayscale, however;
• Images with low signals are noisy and
• Images with high signal are associated with high dose
Exposure Indicators describe
image quality in
terms of the signal to
noise ratio
(SNR)
Underexposed, low SNR Overexposed?, high SNR
18
35
Exposure Indicators
CR and DR systems assess the recorded signal to indicate whether the radiographic technique used is appropriate
• Tests with defined beam conditions are used to verify that correct indicators are being reported
• Recommended exposure indicator ranges are used by technologists to check each radiographic exposure
36
Region to assess signal indicator
Systems vary in the region used to assess the signal for an image.
• Full Image
• Regular regions
• Anatomic regions
J. Shepard and M. Flynn presentation
19
37
Region to assess signal indicator IEC 62494-1
• Gray histogram for the entire image
• Black histogram for the anatomic region (relevant region)
38
Computation of an exposure indicator
…. computed from the probability distribution of signal values in the relevant image region, based on the median
Manufacturers have adopted proprietary methods
• Algorithms, values, and calibration methods are widely different, leading to confusion amongst users
• Inappropriate image segmentation or histogram ‘values of interest’ range can produce inaccuracies
20
39
Summary of manufacturer Exposure Indices Manu-
facturer Indicator
Name Symbol Units
Exposure Dependence
Calibration Conditions
Fujifilm S Value S Unitless 200/S X (mR) 80 kVp, 3 mm Al “total
filtration” S=200 @ 1 mR
Kodak Exposure Index EI mbels EI + 300 = 2X 80 kVp + 1.0 mm Al + 0.5
mm Cu EI = 2000 @ 1 mR
Agfa Log of Median of histogram
lgM bels lgM + 0.3 = 2X for 400 Speed Class, 75
kVp + 1.5 mm Cu lgM=1.96 at 2.5 µGy
Konica Sensitivity Number
S Unitless for QR = k,
200/S X (mR) for QR=200, 80 kVP S=200
@ 1 mR
Canon Reached
Exposure Value REX Unitless
Brightness = c1, Contrast = c2,
REX X 1
for Brightness = 16, Contrast = 10,
REX ≈ 106 @ 1 mR1
Canon EXP EXP Unitless EXP X
80 kVp, 26 mm Al HVL = 8.2 mm Al DFEI = 1.5
EXP = 2000 @ 1 mR 1 From empirical data
40
Summary of manufacturer Exposure Indices Manu-
facturer Indicator
Name Symbol Units
Exposure Dependence
Calibration Conditions
GE Uncompensated
Detector Exposure
UDExp mGy Air KERMA UDExp X (μGy) 80 kVp, standard filtration,
no grid
GE Compensated
Detector Exposure
CDExp mGy Air KERMA CDExp X (μGy)
GE Detector Exposure Index DEI Unitless DEI ≈ 2.4X (mR) 1 Not available
Swissray Dose Indicator DI Unitless Not available Not available
Imaging Dynamics Company
Accutech f# Unitless 2f#=X(mR)/Xtgt(mR) 80 kVp + 1 mm Cu
Philips Exposure Index EI Unitless 100/S X (mR) RQA5, 70 kV, +21 mm Al, HVL=7.1 mm Al
Siemens Medical Systems
Exposure Index EXI mGy Air KERMA X(mGy)=EI/100 RQA5, 70 kV +0.6 mm Cu,
HVL=6.8 mm Al
Alara CR Exposure Indicator Value EIV mbels EIV + 300 = 2X 1 mR at RQA5, 70 kV, +21 mm Al,
HVL=7.1 mm Al => EIV=2000
iCRco Exposure Index none Unitless Exposure Index log [X (mR)] 1 mR at 80 kVp + 1.5 mm Cu => =0
21
41
Manufacturer Symbol 50 mGy 100 mGy 200 mGy
Canon (Brightness =16, contrast = 10
REX 50 100 200
IDC (ST = 200) F# -1 0 1
Philips EI 200 100 50
Fuji, Konica S 400 200 100
Kodak (CR, STD) EI 1700 2000 2300
Siemens EI 500 1000 2000
Approximate EI Values vs. Receptor Exposure
….. The need for a standard clearly evident
42
Standardization
• American Association of Physicists in Medicine Task Group 116 and International Electrotechnical Commission (IEC)
• Collaborative effort • Physicists • Manufacturers/Vendors representatives • MITA (Medical Imaging and Technology Alliance)
• Develop common “Exposure Indices” and “Deviation Indices” across detectors and manufacturers/vendors
• Provide means for placing data in DICOM metadata
22
43
AAPM TG 116 The AAPM TG 116 report on exposure indicators was published in July of 2009
IEC Standard IEC published a standard for Exposure Index definitions in August of 2008
44
Description of Exposure Indices Parameters
AAPM
TG116 Med Physics 2009
IEC
62494-1 IEC:2008
Exposure Index
Air-kerma at the receptor
KIND = KCAL (μGy)
EI = KCAL × 100 μGy-1
(unitless)
Calibration Energy
RQA-5
66 - 74 kVp
RQA-5
66 - 74 kVp
Calibration Filtration
RQA-5 Equivalent
0.5 mm Cu (+ 0-3 mm Al) or 21 mm Al
6.8 ± 0.2 mm Al HVL
RQA-5 Equivalent
0.5 mm Cu + 2 mm Al or 21 mm Al
6.8 ± 0.3 mm Al HVL
Deviation Index
Deviation Index
DI = 10*log10(KIND/KTGT)
Deviation Index
DI = 10*log10(EI/ET)
DI format Signed decimal string with
1 decimal point Unspecified
23
45
Deviation Index (DI)
• EIT is a target index value that is to be determined for each
body part b, view v, procedure type, and clinical site
• When EI equals EIT, DI = 0 • DI = +3.0 for 2x target exposures • DI = -3.0 for ½ target exposure • ± 1 is one step on a standard generator mAs control or
AEC compensation (ISO R5 scale)
Exposure Indices
1010( . )T
EIDI Log
EI b v
46
Num
be
r o
f p
ixe
ls
Pixel Value
Values of Interest
EI = EIT
DI = 0.0 EI and DI calculated from
this pixel value
What about VOI modification by the technologist?
Need to have robust methods of determining DI
24
47
Num
be
r o
f p
ixe
ls
Pixel Value
VOI recognition algorithm fails
• Gonadal shields, prosthetics, etc.
• False DI reported
Correct
Values of Interest
EI = EIT
DI = 0.0
EI and DI calculated from
this pixel value
Incorrect Values of Interest
EI and DI incorrectly calculated
from this pixel value
EI EIT
DI = -1.3
Need robust methods of determining EI & DI
48
Need robust methods of determining EI & DI
Num
be
r o
f p
ixe
ls
Pixel Value
User adjusts VOI for proper grayscale rendition manually,
and DI returns to zero
Incorrect Values of Interest
EI and DI incorrectly calculated
from this pixel value EI EITGT
DI = -1.3
Correct
Values of Interest
EI and DI calculated from
this pixel value
EI = EITGT
DI = 0.0
25
49
Need to determine recommendations for repeats
• DI target is -2.0 to +2.0
• Check for noise. Consult with radiologist on need for repeat if EI is 63% of target (DI -2)
• Investigate cause (do not repeat) if EI is between 160% and 200% of target (+2.0 DI +3.0)
• Consult with radiologist (check for saturation) on need for repeat and counsel of technologist if EI is 200% of target (DI +3.0)
50
IEC 62494-1: Target Exposure Index EIT
• EIT may depend on detector type, examination type, diagnostic question and other parameters
• Establishing target exposure index values needs medical knowledge – may be done by professional societies
• EIT values should be provided as a data base in the digital imaging system
Ulrich Neitzel Project Leader,
Convenor IEC SC62B WG 43
26
51
Caveats
• The EI does not describe patient dose
• EI is derived from detector signal (dose at the detector)
• The EI is not a dose measurement tool
• Dose calibration only valid at one radiation quality
• Images with same EI obtained on different digital systems might not have similar image quality
• Influence of detector DQE, scattered radiation, beam quality differences
Ulrich Neitzel Project Leader,
Convenor IEC SC62B WG 43
52
Exposure Index & Deviation Index monitoring
• Collect EI and DI for every image and analyze • By technologist • Technique factors • X-ray system • Plate scanning unit (CR) • Processing unit (CR - DR) • Anatomical view
• Longitudinal studies • Track performance over time • Mean and Standard Deviation of EI and DI • Watch for trends upward (Dose Creep)
27
• RIS-PACS Integration – Data Synchronization, Validation
– Interpretation & Results Reporting
• Modality Worklist (MWL) – RIS-driven data transfer of exam information
– Aid the technologist for protocol and room setup
• Image Distribution (The Internet) – Clinical Review, OR, Patients, Conferences
– Enterprise Integration - EMR
– Teleradiology
Image Integration and Distribution
• TCP/IP – Standard Communications Protocol
– The Internet
• HL7 – Health Level 7
– RIS / HIS
• DICOM 3.0 – Digital Imaging COmmunications in Medicine v3.0
– PACS
• HTTP – Hyper-Text Transport Protocol
– The World Wide Web
Communication Protocols
28
A typical imaging solution
Typical PACS workstation configuration
•1536x2048
•10 bit
•Grayscale or
color
•400 Cd/m2
Navigation
Monitor(s)
1920x1080
color
Digital voice
dictation
support
29
Basic Workstation Software Reqt’s
• Filters: sort studies by modality, location, time, etc.
• Worklist functionality: automate workflow
• Hanging protocols: arrange images and display
• Retrieving priors: pre-fetch or all spinning disk
• Graphic user interface: tool palette parameters
• Mechanical interface: keyboard, mouse, other
• User preferences: individual preferences for above
• American College of Radiology
• National Electrical Manufacturers Association
(NEMA)
• Established 1983, first published 1985
• Followup standards in 1988 (ACR-NEMA 2.0)
• 1993, DICOM 3.0 published and continuously
updated
30
What does DICOM do?
• Addresses 5 areas of functionality
– Transmission and persistence of complete objects
(images, waveforms, documents)
– Query & Retrieval of such objects
– Performance of specific actions (e.g. film printing)
– Workflow management (support of worklists)
– Quality and consistency of image appearance
(both display and print)
• Network configurations
– Application Entity (AE) title, IP address, TCP/IP port
number
What does DICOM do?
• DICOM storage
– e.g., CT image storage SOP class, CR image storage SOP class
• DICOM print
– Basic grayscale print management SOP class (SCU only)
• Query / Retrieve
– Poll a DICOM device for a list of studies or patients, then retrieve
one or more
• Worklist Management
– Download a list of “scheduled procedures” to the modality from
the RIS through a worklist management provider (PACS Broker)
• Modality Performed Procedure Step (MPPS)
– Modality tells RIS that the procedure has been performed
31
DICOM Metadata……
PACS capability & needs
• Hanging protocols—radiologist specific
• API – application program interface
• 3rd party add-ons
• Presentation State
• VOI LUT (Value Of Interest Look-up-table)
• De-identification / anonymization (HIPAA)
• Part 14 DICOM export
• ……… and so on…..
32
IMAGE DISTRIBUTION:
Web and the Internet and Teleradiology
Image
Archive
(TB)
Control
Software
Review
Station
Review
Station
Web
Servers
Hospital
PCs
Radiology Departments
US CR
Clinic
PCs
Office
PCs
Home
PCs
Radiologists
Remote
Radiologists
Imaging
Centers CT MRI
Courtesy of Dr. Keith Dreyer
PACS installation planning
• Location of server and major hardware
• PC requirements for workstations,
environment and power considerations
• “Mini-PACS” for ultrasound, nuclear medicine
– Color monitors
– Application specific workstation requirements
• Reading room
– Lighting
– Furniture
33
Challenges
• Integration of different systems
• IHE: Integrating the Healthcare Enterprise
– Addresses INTEROPERABILITY of systems
– Provides INTEGRATION PROFILES and a
framework for performing needed functionality and
workflow
• http://www.IHE.org
PACS quality control
• Interfaces
• Redundancy and emergency backup
• Software verification (distance accuracy
measurements, quantitative measurements)
• MONITORS
• Verification of correct data
• Display and viewing conditions
• Image compression & archiving
• Disaster Recovery and backup plans
34
SUMMARY Enterprise distribution of images is crucial
for implementation and application of technology
• Cassetteless, active detector radiography devices (flat panel) are becoming the detectors of choice
• Proliferation of web-based PACS and unified patient database (instead of Radiology centric orientation) is here
• New opportunities
– Image acquisition and image processing tools
– Imaging technology innovation for diagnosis and intervention