157
Quality Assurance in Mammography Melissa C. Martin, M.S., FACR [email protected] 310-612-8127 Sacramento Mammograpy Society Meeting February 9, 2008

Melissa C. Martin, M.S., FACR

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Melissa C. Martin, M.S., FACR

Quality Assurance in Mammography

Melissa C. Martin, M.S., FACR

[email protected]

Sacramento Mammograpy Society Meeting

February 9, 2008

Page 2: Melissa C. Martin, M.S., FACR

The Evolution of Mammography

Page 3: Melissa C. Martin, M.S., FACR

What is Mammography?

• 2-D X-ray breast imaging

Page 4: Melissa C. Martin, M.S., FACR

History of Mammography

1930:Pre compression

Page 5: Melissa C. Martin, M.S., FACR

History of Mammography

1951:Early method ofcompression

Page 6: Melissa C. Martin, M.S., FACR

History of Mammography

1966:First dedicatedmammographysystem

Page 7: Melissa C. Martin, M.S., FACR

Screen-Film Mammography

For 35+ years, screen-film has been the “gold standard” for breast cancer detection

Page 8: Melissa C. Martin, M.S., FACR

Screen Film MammographyX-ray source

Screen-film cassette

Film Processor Film viewer

Page 9: Melissa C. Martin, M.S., FACR

State of the Art Mammography

2000: Digital mammography

Page 10: Melissa C. Martin, M.S., FACR

X-ray source

Digital detector

Full Field Digital Mammography

Acquisition computer

Softcopy display

Page 11: Melissa C. Martin, M.S., FACR

System Quality Control

• Pre-Acquisition• Acquisition• Image Processing and Display• Storage/Archive

The system is only as good as the weakest link-more than just the modality itself

Page 12: Melissa C. Martin, M.S., FACR

QC Recommendations

• Daily QC checks (technologist)– System warmup

• Detector uniformity• Spatial resolution• Contrast resolution (signal to noise ratio)• Geometric accuracy• Clinical review and artifact identification

– Logging / preventive maintenance

Page 13: Melissa C. Martin, M.S., FACR

QC Recommendations

• Each image: QC verification (technologist)– Patient positioning– Image orientation– Image markers– Processing and grayscale rendering– Artifacts (dropouts, lines, scratches, etc.)– Exposure Index- Patient Dose Estimates

Page 14: Melissa C. Martin, M.S., FACR

FDA Approved Full-Field Digital Mammography Systems

GE Senographe 2000D- January 28, 2000

Fischer SenoScan– September 25, 2001

Lorad Selenia– October 2, 2002

GE Senographe DS– February 19, 2004

Siemens MammomatNovation DR- August 20, 2004

GE Senographe Essential - April 11, 2006

Fuji FCRm- July 10, 2006

Page 15: Melissa C. Martin, M.S., FACR

GE 2000DGE 2000D

System Components

Page 16: Melissa C. Martin, M.S., FACR

GE Seno DS & EssentialGE Seno DS & Essential

Page 17: Melissa C. Martin, M.S., FACR

LoradLorad SeleniaSelenia

Page 18: Melissa C. Martin, M.S., FACR

SiemensSiemensMammomatMammomat NovationNovationDRDR

Page 19: Melissa C. Martin, M.S., FACR

Fischer Fischer SenoscanSenoscan

Page 20: Melissa C. Martin, M.S., FACR

MQSA National Statistics

• Certified facilities as of 2/1/08: 8861• Certified facilities as of 10/1/2006: 8832• Total accredited units (2/1/08): 13,575• Certified facilities with FFDM units: 2743

(2/1/08) - 31% of facilities• Accredited FFDM units (2/1/08): 4132

- 30.4% of all units

Page 21: Melissa C. Martin, M.S., FACR

MQSA National Statistics (Butler/Wilcox)

Page 22: Melissa C. Martin, M.S., FACR

MQSA National Statistics (Butler/Wilcox)

Page 23: Melissa C. Martin, M.S., FACR

Why Full Field Digital Mammography?Why Full Field Digital Mammography?

• Technical Reasons

• Clinical Reasons

• Practical Reasons

• Technical Reasons

• Clinical Reasons

• Practical Reasons

Page 24: Melissa C. Martin, M.S., FACR

Why FFDM - Technical ReasonsWhy FFDM - Technical Reasons• Wider dynamic range

– 3-4 times higher• Linear detector response

– Over full range• Better low contrast resolution

– Enhanced visualization of masses• Higher DQE

– Higher absorption efficiency– Dose or SNR– Important for detector as well as entire system

• Wider dynamic range– 3-4 times higher

• Linear detector response– Over full range

• Better low contrast resolution– Enhanced visualization of masses

• Higher DQE– Higher absorption efficiency– Dose or SNR– Important for detector as well as entire system

Page 25: Melissa C. Martin, M.S., FACR

Technical Reasons

Page 26: Melissa C. Martin, M.S., FACR

Why FFDM- Clinical Reasons

Page 27: Melissa C. Martin, M.S., FACR

DMIST Results

Page 28: Melissa C. Martin, M.S., FACR

Why FFDM- Practical Reasons

Page 29: Melissa C. Martin, M.S., FACR

Weaknesses of Digital Mammography

Page 30: Melissa C. Martin, M.S., FACR

Dose Creep

Page 31: Melissa C. Martin, M.S., FACR

31Butler/wilcox - RSNA 2006

Page 32: Melissa C. Martin, M.S., FACR

Fundamentals of Digital Mammography• Theory of digital detectors• Pixel size considerations• Anti-scatter grids• Gantry system components• Acquisition system components• Softcopy workstation components• Computer aided diagnosis• Advanced applications

Page 33: Melissa C. Martin, M.S., FACR

Digital Mammography Detector Technical Background

Page 34: Melissa C. Martin, M.S., FACR

Direct Vs. Indirect Conversion

a-Selenium Semiconductor

TFT array

---+++

Selenia Direct Digital

CsI Scintillator

TFT array or CCD

Indirect DigitalX-ray photon X-ray photon

Line spread functions

Film

X-ray photon

Film

Phosphor

Page 35: Melissa C. Martin, M.S., FACR

Available FFDM Systems

direct conversion TFT

MammomatNovation (DR)

•Siemens Medical

direct conversion TFT

Selenia•Hologic/Lorad

indirect conversion, slot scanning

Senoscan•Fischer Imaging(No longer available as new)

Indirect conversion TFT

Senographe 2000Dand Essential

•GE Medical Systems

Page 36: Melissa C. Martin, M.S., FACR

Image Sharpness

Modulation Transfer Function (MTF)measures resolution∗ Higher MTF = Sharper images∗ Higher MTF = Superior detail image

contrast

Page 37: Melissa C. Martin, M.S., FACR

0

0.2

0.4

0.6

0.8

1

0 2 4 6 8 10

Spatial Frequency (lp/mm)

Pres

ampl

ed M

TF

Modulation Transfer Function

Kodak Min-R 2000

GE Senographe 2000D

Lorad Selenia

Fischer Senoscan

* Data From PMA Submissions, European Journal Radiology

Page 38: Melissa C. Martin, M.S., FACR

What Does Higher MTF Mean?• Higher MTF means increased image

sharpness, less blur• Sharper images of small details means

better detection• Better detection means fewer missed

cancers• Sharper microcalcifications means better

visualization of morphology & higher confidence in diagnosis

Page 39: Melissa C. Martin, M.S., FACR

Dose EfficiencyDetective Quantum Efficiency (DQE) measures dose efficiency

∗ Higher DQE = Superior image quality

∗ Higher DQE = Reduced patient dose

Page 40: Melissa C. Martin, M.S., FACR

0%

10%

20%

30%

40%

50%

60%

70%

0 2 4 6 8 10 12

Spatial Frequency (lp/mm)

DQ

EDetective Quantum Efficiency

Kodak Min-R 2000

GE Senographe 2000D

Lorad Selenia

Fischer Senoscan

* Data From PMA Submissions

Page 41: Melissa C. Martin, M.S., FACR

What Does Better DQE Mean?

• Better DQE means less noise in the image• Less noise means better detection of low

contrast pathology, e.g., masses• Better detection means fewer missed

cancers• Lower noise means better visualization &

higher confidence in diagnosis• Better DQE means flexibility to reduce dose

Page 42: Melissa C. Martin, M.S., FACR

a-Se Direct Capture Detector*

• 70 μm TFT pixel pitch • 24 cm x 29 cm active

image area

*Currently used by Hologicand Siemens

Page 43: Melissa C. Martin, M.S., FACR

Pixel Size Considerations• Need to optimize conflicting requirements: Smaller

pixels give higher resolution, but files are larger. Monitors can’t display large FFDM images in their entirety.

• If resolution is too low, it may impair accurately imaging microcalcifications at earliest stage of formation (~ 150 μm).

• In direct conversion systems, smaller pixel sizes provides higher resolution. This may not be true for indirect conversion systems, which are scintillator limited.

Page 44: Melissa C. Martin, M.S., FACR

System ComponentsSystem Components

Page 45: Melissa C. Martin, M.S., FACR

Selenia Gantry

• Built on analog Lorad M-IV design

• Optimized for digital imaging

Page 46: Melissa C. Martin, M.S., FACR

Selenia High Transmission Cellular (HTC®) Grid

• Integrated into the digital detector • Cross-hatch design reduces scatter in

X & Y direction• Allows 70-80% transmission of

primary beam• Automatic retraction mechanism

facilitates geometric magnification exams

Page 47: Melissa C. Martin, M.S., FACR

Image Receptor Size

24 x 29 cm Detector

18 x 24 cm Detector

• Unlike film, you only get one size per system!

Page 48: Melissa C. Martin, M.S., FACR

With 18 x 24 cm Detector Size:• >30% of women will undergo

– Double exposure of overlapping tissue– Multiple additional compressions– Extended exam times

• Radiologists will have to review a total of 7 to 8 images per exam

With 24 x 29 cm Detector Size:• Need to accommodate imaging of smaller breasts

Page 49: Melissa C. Martin, M.S., FACR

Smart Paddle System

• Microprocessor in each Smart Paddle controls movement of collimation

• When paddle is shifted, collimation automaticallymoves to proper position

CC

LMLO

RMLO

Page 50: Melissa C. Martin, M.S., FACR

• Integrated into Smart Paddle system

• Provides proper & consistent compression across the entire breast– Reduced motion artifacts– More uniform compression– Maximum patient comfort

Fully Automatic Self-adjusting Tilt (FAST) Paddle

Page 51: Melissa C. Martin, M.S., FACR

Bi-Angular X-ray TubeAllows higher mA loading with smaller effective focal spot size for improved resolution, especially on magnification views

Page 52: Melissa C. Martin, M.S., FACR

• Two modes of operation– Virtual multi-

sensor (fully automatic)

– 7 User-selectable positions

• 12.5 cm travel from chest wall– Accommodates

imaging of larger breasts

Automatic Exposure Control (AEC)

Page 53: Melissa C. Martin, M.S., FACR

Fully Automatic Collimation• Easier & quicker workflow• Easier positioning on

magnification & spot compression views

• System selections can be overridden easily

• Collimation follows compression paddle

• Stored image size reduces with smaller paddle

Page 54: Melissa C. Martin, M.S., FACR

Auto FiltrationMolybdenum or Rhodium is automatically selected for reduced dose & increased penetration of dense tissue

Page 55: Melissa C. Martin, M.S., FACR

Selenia Image Acquisition Station• Integrated X-ray control • Familiar M-IV interfaces• Short-term archive• Built-in QC & service tools• DICOM 3.0 connectivity

• Flat panel display monitor

• Automatic MQSA repeat analysis report

• Small foot print

Page 56: Melissa C. Martin, M.S., FACR

Flexible Technology InterfacesFlexible Technology Interfaces

• Fully integrated, vendor independent Display of CAD markers & results

• Compatibility with multiple vendors of reporting software

• Multiple approved DICOM compatible printers

• Fully integrated, vendor independent Display of CAD markers & results

• Compatibility with multiple vendors of reporting software

• Multiple approved DICOM compatible printers

Page 57: Melissa C. Martin, M.S., FACR

There Are Currently Several FDAThere Are Currently Several FDA--Approved Laser Imagers for Digital Approved Laser Imagers for Digital MammographyMammography

•• Agfa DS4500MAgfa DS4500M

•• Kodak 8600 Laser ImagerKodak 8600 Laser Imager

•• Kodak 8610 Laser ImagerKodak 8610 Laser Imager

•• Kodak 8900MKodak 8900M

•• Fuji Fuji DrypixDrypix 7000 & 5000 7000 & 5000

•• Fuji Fuji DrypixDrypix FMFM--DP LDP L

•• Konica Konica DryProDryPro 793793

Page 58: Melissa C. Martin, M.S., FACR

OD Requirements for OD Requirements for HiHi--Resolution Laser ImagersResolution Laser Imagers

DDmaxmax > 3.5 OD> 3.5 OD

MidMid--density > 1.5 ODdensity > 1.5 OD

Page 59: Melissa C. Martin, M.S., FACR

Kodak daily sensitometryKodak daily sensitometry

Laser Processor QCLaser Processor QC

Base + FogBase + Fog

Density Difference Density Difference –– OD closest to 2.20 OD closest to 2.20 minus OD closest but not less than 0.45minus OD closest but not less than 0.45

MidMid--density density –– step closest to but not less step closest to but not less than 1.20than 1.20

DDmaxmax

Action Limits:Action Limits:

MD & DD MD & DD ++ 0.15 OD0.15 OD

B+Fog = 0.03B+Fog = 0.03

DDmaxmax ++ 0.250.25

Page 60: Melissa C. Martin, M.S., FACR

Selenia Softcopy Workstation

Page 61: Melissa C. Martin, M.S., FACR

Softcopy Image Review StationSoftcopy Image Review Station• Dual 5 Megapixel displays• Display 10-bit grayscale • Configurable doctor-

specific hanging protocols• Supports DICOM For-

Presentation and For-Processing

• Display For-Presentation images from other vendors and other modalities

• Supports double reading• Standard workstation

tools: window, zoom, annotation, measurement, …

Page 62: Melissa C. Martin, M.S., FACR

Soft Copy Workstation ControlsSoft Copy Workstation Controls• User configurable workflow

– Hanging protocol and workflow defined by individual users

– Accommodates multiple users• Custom designed workflow keypad

– Places all controls at physician’s fingertips

– Allows attention to focus solelyon case review

• User configurable workflow– Hanging protocol and workflow

defined by individual users– Accommodates multiple users

• Custom designed workflow keypad– Places all controls at physician’s

fingertips– Allows attention to focus solely

on case review

Page 63: Melissa C. Martin, M.S., FACR

High Volume ThroughputHigh Volume Throughput• Supports Screening &

Diagnostic Mammography• Sub-second delay between

cases• One softcopy workstation

supports up to 4 Selenia Acquisition Stations

• Supports optional automatic double-reading

• Supports Screening & Diagnostic Mammography

• Sub-second delay between cases

• One softcopy workstation supports up to 4 Selenia Acquisition Stations

• Supports optional automatic double-reading

Page 64: Melissa C. Martin, M.S., FACR

Proprietary Image Processing

• Automatic image processing – Optimally enhances

each digital mammogram

• Mammography specific algorithms– Designed specifically

for direct capture Selenia images

• Automatic image processing – Optimally enhances

each digital mammogram

• Mammography specific algorithms– Designed specifically

for direct capture Selenia images

• Peripheral Contrast Enhancement (PCE)– Optimal skin line

visibility– Optimal image

contrast applied across entire breast for different tissue densities

Page 65: Melissa C. Martin, M.S., FACR

Peripheral Contrast Enhancement (PCE)Peripheral Contrast Enhancement (PCE)

PCE Off PCE Applied

Page 66: Melissa C. Martin, M.S., FACR

Thickness CompensationThickness Compensation

Goal: Visualize the breast from chest wall to skin line.Algorithm: 1. Estimate the thickness of the breast at all pixels2. Add a nearly-equivalent thickness [of water] to raise gray level. Original TC Processed

*Note: This is only applied near the skin line

Page 67: Melissa C. Martin, M.S., FACR

Thickness Compensation

Page 68: Melissa C. Martin, M.S., FACR

Digital Mammography Workflow

Page 69: Melissa C. Martin, M.S., FACR

Technologist Workstation• TWS is the means to get the digital image to every

place you would take a film to view in the analog world.

• TWS will be used to assist with workflow in the mammography suite. – TWS used for viewing priors, in the tech area– Annotations done by the radiologist will be sent from

the SCW to the TWS for review– A notification will alert the technologist when

annotated images arrive.

Page 70: Melissa C. Martin, M.S., FACR

Image Management

• Digital mammography images require extensive archiving space & increased bandwidth– Single system can generate 1 Terabyte of Data/year

(50 patients/day)– 2 – 3 Terabytes With Images From Prior Years

• Many PACS may be negatively impacted by demands of digital mammography– Insufficient storage space– Inadequate bandwidth to support acceptable

transmission speed

Page 71: Melissa C. Martin, M.S., FACR

FFDM Mobile Applications

Network Connectivity to Hospital PACS

On-board Image Storage

Satellite Connection toHospital PACS

Page 72: Melissa C. Martin, M.S., FACR

QC - Part 2 - SAMS Meeting 2/08

• QC Procedures

Page 73: Melissa C. Martin, M.S., FACR

FFDM Quality Control

• MQSA Final Regulation– Screen-Film QA/QC– Manufacturer QA/QC for FFDM System

• Manufacturer’s FFDM QC Manual

Page 74: Melissa C. Martin, M.S., FACR

LORAD Selenia QC TestsQC Manual tests:• Medical Physicist

– Annually• Radiologic Technologist

– Daily– Weekly– Biweekly– Monthly– Quarterly– Semiannually

Page 75: Melissa C. Martin, M.S., FACR

Medical Physics Annual Tests• Mammographic Unit Assembly Evaluation• Collimation Assessment• Artifact Evaluation• kVp Accuracy and Reproducibility• HVL Measurement• Evaluation of System Resolution• Breast Entrance Exposure and Glandular Average Dose• Radiation Output Rate• Phantom Image Quality Evaluation• Signal-To-Noise and Contrast-To-Noise Measurements• Viewbox Luminance and Room Illuminance• Softcopy Workstation QC

Page 76: Melissa C. Martin, M.S., FACR

QC Tests for FFDM Units:Tests and Frequencies Differ

Not requiredBi-weeklyWeeklyWeeklyLorad

MonthlyWeeklyWeeklyDailyFischer

MonthlyWeeklyMonthlyMonthlyGE

MTF/ Sys Res

Flat FieldSNRSoft Copy Display

Test

Ex: Tech Test Frequencies

Page 77: Melissa C. Martin, M.S., FACR

LORAD Selenia FFDM System

• RT QC Tests

Page 78: Melissa C. Martin, M.S., FACR

Technologist QC Tests

• Laser Printer Quality Control (Daily)• SNR and CNR Measurements (Weekly)• Softcopy Workstation QC (Weekly)• Phantom Image (Weekly)

Page 79: Melissa C. Martin, M.S., FACR

Laser Printer Quality Control

• SMPTE Pattern from Acquisition Workstation (Daily)

• Measure 10%, 40% and 90% Density Patches• Daily Performance

– 40% density tracks within ±15%– (40% - 10%) density tracks within ±15%– 90% density tracks within ±15%

• NOTE: Printer uses linear LUT

Page 80: Melissa C. Martin, M.S., FACR

SNR and CNR Measurements• ACR Accreditation Phantom with Disc

(Weekly)• Measure SNR and CNR from Image

Page 81: Melissa C. Martin, M.S., FACR

SNR and CNR Measurements

• Passing Criteria– SNR at least equal or greater than 40– CNR should stay within ±15% of

measurement obtained during acceptance testing of system

Page 82: Melissa C. Martin, M.S., FACR

LORAD Selenia FFDM System

• Workstation QC

Page 83: Melissa C. Martin, M.S., FACR

Room Illuminance

• From the ACR QC Manual, 1999, the room illuminance must be below 50 lux.

• From the February 2006 version of the Medical Physicist Equipment Evaluation forms from ACR, room illuminance for soft copy displays must be below 20 lux.

Page 84: Melissa C. Martin, M.S., FACR

Softcopy Workstation QC

• Performed Weekly– Black level measurement– White level measurement

• Performed Monthly– LUT conformance

• Performed Quarterly– White field uniformity

Page 85: Melissa C. Martin, M.S., FACR

Softcopy Workstation QC

• Data Logged by Software• Warning Flags and Errors if Conformance

Fails

Page 86: Melissa C. Martin, M.S., FACR

Phantom Image

• Review Weekly Phantom Image Results– Measured film densities– Phantom objects seen (printer, monitors)

• 5 fibers• 4 speck groups• 4 masses

Page 87: Melissa C. Martin, M.S., FACR

LORAD Selenia FFDM System

• FDA Conformance Regulations

Page 88: Melissa C. Martin, M.S., FACR

Conformance• Federal regulations, 900.12(e), specify that any

time quality control test results fall outside of the action limits for systems with image receptor modalities other than screen-film, the source of the problem shall be identified and corrective action shall be taken before any further examinations are performed.

• Full Field Digital Mammography systems are affected by this ruling

Page 89: Melissa C. Martin, M.S., FACR

QC Alternative Standard• The majority of the required QC testing performed on

a FFDM system applies to system components other than the digital image receptor.

• Only a small portion of the quality control testing performed on an FFDM system is specific to the digital image receptor.

• Thus, alternative standards has been approved by FDA on the QC testing of several of the FFDM systems.

Page 90: Melissa C. Martin, M.S., FACR

QC Alternative Standards for FFDM Units• Three different corrective action

categories– Action Category A– Action Category B– Action Category C

Page 91: Melissa C. Martin, M.S., FACR

Action Category A

• Applies to performance testing of the digital image receptor

• Corrective action shall be taken before any further examinations are performed– Evaluation of System Resolution– Breast Entrance Exposure and Average Glandular Dose– Phantom Image Quality Evaluation– SNR and CNR Measurements

Page 92: Melissa C. Martin, M.S., FACR

Action Category B• Applies to performance testing of diagnostic devices

used for mammographic image interpretation• Corrective action shall be taken before the device can

be used for mammographic image interpretation• Clinical imaging can be continued and alternative

approved diagnostic devices shall be used for mammographic image interpretation– Phantom Image Quality Evaluation– Softcopy Workstation QC– Laser Printer Quality Control

Page 93: Melissa C. Martin, M.S., FACR

Action Category C

• Applies to performance testing of rest of the system• Corrective action shall be taken within thirty days of

the test date• Clinical imaging and mammographic image

interpretation can be continued during this period– All other tests

Page 94: Melissa C. Martin, M.S., FACR
Page 95: Melissa C. Martin, M.S., FACR
Page 96: Melissa C. Martin, M.S., FACR
Page 97: Melissa C. Martin, M.S., FACR
Page 98: Melissa C. Martin, M.S., FACR
Page 99: Melissa C. Martin, M.S., FACR
Page 100: Melissa C. Martin, M.S., FACR
Page 101: Melissa C. Martin, M.S., FACR
Page 102: Melissa C. Martin, M.S., FACR
Page 103: Melissa C. Martin, M.S., FACR
Page 104: Melissa C. Martin, M.S., FACR
Page 105: Melissa C. Martin, M.S., FACR
Page 106: Melissa C. Martin, M.S., FACR
Page 107: Melissa C. Martin, M.S., FACR
Page 108: Melissa C. Martin, M.S., FACR
Page 109: Melissa C. Martin, M.S., FACR
Page 110: Melissa C. Martin, M.S., FACR
Page 111: Melissa C. Martin, M.S., FACR
Page 112: Melissa C. Martin, M.S., FACR
Page 113: Melissa C. Martin, M.S., FACR
Page 114: Melissa C. Martin, M.S., FACR
Page 115: Melissa C. Martin, M.S., FACR

Siemens Novation

Page 116: Melissa C. Martin, M.S., FACR
Page 117: Melissa C. Martin, M.S., FACR
Page 118: Melissa C. Martin, M.S., FACR
Page 119: Melissa C. Martin, M.S., FACR
Page 120: Melissa C. Martin, M.S., FACR

Phantom Image Quality Check

Page 121: Melissa C. Martin, M.S., FACR

Fuji Digital Mammography QCTechnologist’s Tests

TESTCNR-Contrast/NoisePhantom ImagePrinter/MonitorVisual ChecklistRepeat AnalysisCompressionIP Fog

FREQUENCYWeeklyWeekly

Per ManufacturerMonthly

QuarterlySemi-annualSemi-annual

Page 122: Melissa C. Martin, M.S., FACR

Fuji Digital Mammography QCTechnologist’s Tests

Printer TestsNo film in cassetteImaging Plate should be erased after 8

hours of non-useImaging Plate should be processed or

erased after each exposurePrinters are monitored according to the

manufacturer’s approved QC program

Page 123: Melissa C. Martin, M.S., FACR

Fuji Digital Mammography QCTechnologist’s Tests

Phantom Images TestConfirms image quality using ACR PhantomTools required - QC Cassette & ACR PhantomMethod - AEC Exposure, process IP using the

Physics, ACR MAPP MenuVisually inspect the image. Same scoring method

as MQSACorrection Period - Before any further

examinations are performed

Page 124: Melissa C. Martin, M.S., FACR

Fuji Digital Mammography QCTechnologist’s Tests

CNR (Contrast/Noise Ratio) TestEvaluates Noise and Contrast using a fixed x-

ray exposureTools needed - 4 cm acrylic block & 0.2 mm AlMethod - Fixed manual techniqueCNR Measurement must be within +/- 20%Correction Period - Before any further

examinations are performed

Page 125: Melissa C. Martin, M.S., FACR

Fuji Digital Mammography QCTechnologist’s Tests

Tests in Accordance with ACR/MQSAVisual Checklist MonthlyRepeat Analysis QuarterlyCompression Test Semi-Annual

Same Action Limits and Corrective Action as MQSA

IP Fog replaces Darkroom Fog test - Semi Annual

Page 126: Melissa C. Martin, M.S., FACR

Fuji Digital Mammography QCTechnologist’s Tests

Imaging Plate FogConfirm that the storage conditions of the exposure

room protects the IP cassettes from scatterTools required - IP CassetteMethod - Tape coin to front of cassette and place in

storage location during test exposuresVisually inspect image for visibility of coinCorrection Period - Before any further examinations

are performed

Page 127: Melissa C. Martin, M.S., FACR

Fuji Digital Mammography QCTechnologist’s Tests

Inter-Plate Consistency - As NeededThe Inter-plate Consistency Test replaces the

Screen Speed TestMust be performed when new IPs and/or

cassettes are introducedCan be performed by the QC technologist under

Qualified Medical Physicist oversight

Page 128: Melissa C. Martin, M.S., FACR

Fuji Digital Mammography QCTechnologist’s Tests

Inter-Plate Consistency - As NeededConfirm that Imaging Plates used in the facility

are similar in sensitivity and image qualityTools Required - new and existing IPs and

Cassettes, 4 cm acrylic blockMethod - Expose cassettes using AEC (time)

ModeAcceptable results: mAs +/- 10$, SNR +/- 15%Correction Period - Within 30 days of the test

Page 129: Melissa C. Martin, M.S., FACR

129Butler/Wilcox - RSNA 2006

Page 130: Melissa C. Martin, M.S., FACR

As good or better than film, but its main advantage is:

Provides ideal platform for advanced applicationsTomosynthesis (3D mammography)Contrast enhanced mammography

As good or better than film, but its main advantage is:

Provides ideal platform for advanced applicationsTomosynthesis (3D mammography)Contrast enhanced mammography

Full Field Digital Mammography (FFDM)

Page 131: Melissa C. Martin, M.S., FACR

Future of Mammography

2008:Tomosynthesis3D mammography

Page 132: Melissa C. Martin, M.S., FACR

What is Tomosynthesis?• 3-D X-ray mammography• Helps reduce tissue overlap (structure noise)

2-D image can obscure

lesions

Thin slices from 3-D

reduce tissue overlap

Page 133: Melissa C. Martin, M.S., FACR

Why Tomosynthesis?

• Elimination of overlapping structures and shadows

• See through dense breasts• Clearer lesion margins• Potential for reduced call-

backs• Potential for reduced dose

Page 134: Melissa C. Martin, M.S., FACR

Tomosynthesis Acquisition

• Limited angle tomography • X-ray tube moves in an arc across the breast• Series of low dose images are acquired at different angles• Total dose similar to single view breast exam

Digital detector

Compression plate

Breast

X-ray tube

Reconstructed planes

Page 135: Melissa C. Martin, M.S., FACR

Conventional 2-D Imaging

Incident X-rays

Objects being imaged

2-D imageImages superimposed on image

Page 136: Melissa C. Martin, M.S., FACR

Tomosynthesis Acquisition

Incident X-rays

Objects being imaged

2-D raw data images

Image from multiple angles

Exposure #1 Exposure #6 Exposure #11

Page 137: Melissa C. Martin, M.S., FACR

Tomosynthesis Acquisition

• X-ray tube moves in an arc around the breast• Series of low dose images are acquired at different

angles • Total dose similar to standard breast exam

Digital Detector

Compression PlateBreast

X-ray Tube

Tube motion

Reconstructed planes

Page 138: Melissa C. Martin, M.S., FACR

Visualization of Microcalcifications

Normalmammogram

Tomosynthesis slice

Page 139: Melissa C. Martin, M.S., FACR

3-D Information

Normalmammogramshows 2 calcs, but what is their spatialrelationship?

Tomosynthesis slices show they are separated by17 mm.

Z = 30 mm. Z = 47 mm.

Page 140: Melissa C. Martin, M.S., FACR

Seeing through overlying tissue

Normal mammogram Tomosynthesis slice

Page 141: Melissa C. Martin, M.S., FACR

Improved Visualization

Normal mammogram Tomosynthesis slice

Page 142: Melissa C. Martin, M.S., FACR

3D Digital Mammography Tomosynthesis

• Multiple views reconstructed into 3D image• Helps solve tissue overlap problems• Initial clinical trials underway

* Works-in-progress

Normal mammogram

Tomosynthesis slices

Page 143: Melissa C. Martin, M.S., FACR

Tomosynthesis Reconstruction

Acquiring images of breast at different angles allows reconstruction of 3-D volume

Page 144: Melissa C. Martin, M.S., FACR

Tomosynthesis ReconstructionAppropriate shifting and adding of raw data reinforces objects at specific height.

Additions improve quantum signal/noise

Height 1 Height 2

Page 145: Melissa C. Martin, M.S., FACR

Raw Data

Page 146: Melissa C. Martin, M.S., FACR

Data Formats

Raw data:11-21 low dosemammogramsviewing breast atslightly differentangles

Page 147: Melissa C. Martin, M.S., FACR

Data Formats

Reconstruction:40-60 cross-sectional slices, one per mm ofcompressedbreast

Page 148: Melissa C. Martin, M.S., FACR

TomosynthesisSystem Requirements• Rapid digital detector readout• High Detective Quantum Efficiency (DQE),

low noise detectors ~ each image low dose• Large area detectors, to image the entire

breast• Short scan times (~ seconds) to reduce

patient motion

Page 149: Melissa C. Martin, M.S., FACR

Tomosynthesis Prototype

FFDM Gantry

Compression paddle

Breast support platform

Selenium image receptor

Page 150: Melissa C. Martin, M.S., FACR

Tomosynthesis Dose

• Total dose similar to conventional single view mammography exposure

• Each single view low dose• Reconstruction recovers quantum

signal/noise in final image by adding the exposures

Page 151: Melissa C. Martin, M.S., FACR

Signal to Noise Improvement• Reconstruction recovers quantum signal/noise from

sequence of low dose images• Relevant noise is structure, not quantum• Improved visibility by removal of overlapping tissues

2-Dimage

3-D Tomo image

Page 152: Melissa C. Martin, M.S., FACR

3-D Information

Mammogramshows 2 calcs, but what is their spatialrelationship?

Tomosynthesisslices show they are separated in depth by17 mm

Z = 30 mm Z = 47 mm

Page 153: Melissa C. Martin, M.S., FACR

Tomosynthesis

• Superior low contrast visibility compared to FFDM or film, even at lower doses

• Clinical trial is being done at matched dose to FFDM– for improved clinical efficacy

Page 154: Melissa C. Martin, M.S., FACR

Features of Tomosynthesis• 3-D X-ray mammography• High in-plane resolution

(mammographic resolution)• Thin cross-sectional slices

(typically about 1 mm)• Same radiation dose as 2D mammography• Rapid scans in seconds• Simple modification of FFDM systems

Page 155: Melissa C. Martin, M.S., FACR

Purposes of Tomosynthesis

• Addresses superimposed tissue problem– See small pathologies more reliably– Reduce superimposed tissues mimicking

pathology• Provides accurate 3D localization

– Helpful for biopsy

Page 156: Melissa C. Martin, M.S., FACR

Current Clinical Status

• In clinical trials in the US and elsewhere• Several manufacturers have announced

plans• No system currently FDA approved • May change the standard of care in

mammography!

Page 157: Melissa C. Martin, M.S., FACR

Thank You !!!

Melissa C. Martin, M.S., FACRTherapy Physics Inc.

879 West 190 St., Ste 419Torrance, CA 90248

Office Phone: 310-217-4114e-mail: [email protected]