Update on ACR Digital Mammography QC Manual - Priscilla Butler

4/28/2015

1

Update on ACR Digital Mammography QC Manual

Priscilla F. Butler, M.S.

Medical Physicist and Senior Director, ACR, Reston, VA

(with thanks to Eric Berns, Ph.D.)

Overview

• Phantom Specifications

• QC Manual

• Explain what we feel is important

• Thank you for being here today

4/28/2015

2

Digital Mammography in the US (as of 3/1/15)

• 13,493 units at 8720 facilities

• Over 96% of all units in US are FFDM

*4 accrediting bodies

*Over 30 models and mfrs approved by FDA

4/28/2015

3

Quality Control: What It and Why Is It Important?Primary Purpose

• Reduce exposure to patients and personnel

• Ensure adequate and consistent patient image

quality• Detect and correct for potential problems, before

they impact patient image quality and care

What it’s not:• Not a detailed technical evaluation of a unit• Not a detailed measure of a limits of a unit• Not the optimization of a unit

• MQSA

– Mammography Quality Standards Act

• ACR

– American College of Radiology

QC Mandates

4/28/2015

4

FDA’s Current DM QC Requirements

• Follow latest version of mfr’s QC manual procedures for unit tested– Lorad (Hologic) allows facility to follow any of their

manuals

• Meet mfr’s performance standards

• Failures must be fixed before use on patients

– Most mfrs applied for alternative standards to allow 30 days for some QC tests

ACR’s Current DM QC Requirements

• Same as FDA’s

• Which are the same as the manufacturer’s

• ACR suggests using manufacturer’s data forms

4/28/2015

5

Screen-Film QC

• Single detector technology

• Single processing technology

• First QC manual in 1990

Manufacturer’s DM QC Requirements

• Tests vary for each manufacturer and model

– Some tests same but names different

– Some tests not required by some manufacturers

• Frequencies vary for each manufacturer and model

• Procedures vary for each manufacturer and model

• Pass/fail criteria vary for each manufacturer and model

4/28/2015

6

Film Printers

• FDA recommends only using printers cleared by FDA’s Office of Device Evaluation for FFDM (but may legally use others)

• Facility must have access to a laser printer (either on-site or someplace else)

• Printer must exist and be tested by MP before the facility performs mammography

Monitors and Workstations• FDA regs require facilities to comply with a QA

program substantially the same as that of FFDM mfr(i.e., GE, Lorad, etc.) – Impractical; even impossible since tests are software-

based

• FDA says – If the monitor/workstation has been cleared by FDA’s ODE

for FFDM, the monitor’s QC manual is “substantially the same” and OK to follow

– If monitor was not cleared by FDA ODE for FFDM facilities must follow one by FFDM mfr

• FDA ODE clears monitors/workstations– Over 500 total; ? have been cleared for FFDM

4/28/2015

7

ACR DM QC Manual

VS.

All of the above may vary with QC manual revisions of same manufacturer/model

ACR DM QC Manual Project

• Subcommittee on Quality Assurance– Chair – Eric Berns, PhD – University of Colorado

• Subcommittee – all volunteers (with day jobs) – Clinical representatives

– MITA representatives (equipment and phantom manufacturers)

4/28/2015

8


• Subcommittee Charge:– Design ACR Accreditation Phantom for FFDM

– Write QC Manual for ACR FFDM Mammography Accreditation Program

• Subcommittee Goals:– Standardize all QC tests for all digital manufacturers

– Standardize test frequencies

– Standardize performance criteria


• QC tests:– Will apply to all manufacturers

– Tests come from a variety of sources (MQSA, SFM, ACRIN DMIST, Manufacturer’s QC programs, MITA, subcommittee clinical experience, etc.)

– Clinically relevant

– User friendly

– Eliminate non-productive testing

• Just because you can test something, doesn’t mean you should!

4/28/2015

9

ACR Digital QC Manual - Structure

• Radiologist’s Section

• Radiologic Technologist’s Section

• Medical Physicist’s Section

• Appendices

• **Clinical Image Quality Section (w/Patient Positioning and Compression and Clinical Image Quality Evaluation) will be revised at a later date and posted on ACR website

ACR Digital QC Manual - Tools

• Will be downloadable to computer or tablet

• Future revisions fully described (in new “Revisions” section)

• Sections and references fully linked

• Optional procedures and forms available

• Management forms available to aid organization

• Procedures liberally illustrated

• Forms in PDF and Excel formats– Excel formats contain formula and functionality– Excel forms downloadable from ACR website

4/28/2015

10


• Subcommittee Goals:

– Standardize all QC tests for all digital Mfrs

– Standardize test frequencies

– Standardize performance criteria

What Will Be New?

4/28/2015

11

The ACR DM Phantom Prototype

Phantom Prototype Design Principles

4/28/2015

12



Based on existing ACR Accreditation Phantom

Similar imaging and scoring to current phantom

Build on experience of QC techs and physicists at ~8,700 USfacilities who already know how to use and score the existing phantom (~25,000+ techs)

4/28/2015

13


Can be used on both SFM & FFDM

Total attenuation matched to current SFM phantom– Similar thickness

– Similar total dose

Permits testing ofthe MQSA 3.0 mGy dose limit (single CC view)

Proposed Scoring Changes

Eliminate subtraction for artifacts

Add “Fail” for artifacts

New pass/fail criteria from

– 4,3,3

– To: 2,3,2

– **But, objects are the same (effective) size as

SFM Phantom

4/28/2015

14

Total Thickness = 4.10 + 0.03 cm

31.0 + 0.1 cm

Cover =Nominal 0.3 cm

Screws

19.0 + 0.1 cm

CNR Cavity (0.1 + 0.005 cm Deep)

Depth of CNR Cavity = 0.1 + 0.005 cm

2.0 + 0.05 cm

Centered Left to Right

2.5 + 0.05 cm

Milled out wax insert area

7.0 cm

(+0.04, -0.00 cm)

Compensator = 0.023 cm

Wax = 0.70 cm + 0.02 cm

Milled out wax insert area

13.0 cm (+0.04, -0.00 cm)1.0 + 0.05 cm

ID Tag

ACR Phantom PrototypeAir Gap = 0.027 cm Nominal

Total Insert Depth = 0.75 cm

Total Thickness Under Insert = 3.05 cm

Test object distance from base of wax = 0.35 + 0.10 cm

1.0 + 0.05 cm

1.0 + 0.05 cm

9.5 cm

Tolerances (Insert Well & CNR Cavity)• Wax insert well depth : + 0.005 cm (+ 2 mils).• Wax insert well width and length : + 0.04 / -0.00 cm.• CNR cavity depth : + 0.005 cm (+ 2 mils).• CNR diameter : + 0.05 cm.

12.98 cm (+ 0.00, - 0.04 cm)

6.98 cm

(+ 0.00, - 0.04 cm)2.0 cm

Wax Insert Specifications with Virtual “Placement Grid”

2.0 cm

0.49 cm

0.49 cm

2.0 cm

2.0 cm each0.49 cm

0.49 cm

Notes:

Test objects to be centered on their respective “placement grid” locations.

0.49 cm perimeter around test object “placement grid”.

0.635 cm (1/4 inch) radius on corners of wax insert.

45o + 5o

Fiber

45o + 5o

Fiber

Fiber specifications

Fiber Length = 1.0 cm + 0.1 cm

Fiber Diameter = See Table

Fiber Placement specs

Speck Placement & Specs

1. Specks to be placed at points on star and middle of star

2. Speck Size (spherical) = See Table

3. Center speck placement to be within + 0.1 cm of center of virtual grid

4. Distance from center speck to center of speck on perimeter = 0.5 cm + 0.1 cm

Mass Placement & Specs

1. Mass pre-cut sphere diameter = 5/8 inch

2. Mass placement to be within + 0.1 cm of center of virtual grid

ID TagID Tag

ID Tag Specs

Virtual Box: height = 0.5 cm, length = 1.8 cm

Location of Virtual Box 1.5 cm

4/28/2015

15

Test Object

Fiber DiameterSpeck Diameter(Glass Spheres)

MassThickness

mm mm mm

1 0.89 + 0.05 0.33 + 0.0100 1.00 + 0.05

2 0.75 + 0.03 0.28 + 0.0083 0.75 + 0.05

3 0.61 + 0.03 0.23 + 0.0069 0.50 + 0.05

4 0.54 + 0.03 0.20 + 0.0059 0.38 + 0.04

5 0.40 + 0.03 0.17 + 0.0084 0.25 + 0.03

6 0.30 + 0.03 0.14 + 0.0070 0.20 + 0.02

Wax Insert Test Object Specifications

Test Object

Fibers (mm)

Specks(mm)

Masses (mm)

ACR 156 FFDM ACR 156 FFDM ACR 156 FFDM

1.56

1.12 0.54 2.00

0.89 0.89 0.40 1.00 1.00

0.75 0.75 0.32 0.33 0.75 0.75

0.61 0.28 0.50 0.50

0.54 0.54 0.24 0.23 0.38

0.40 0.40 0.20 0.25 0.25

0.30 0.16 0.17 0.20

0.14

Summary of Test Object “Visual Equivalency”

4/28/2015

16

Test Object

Fibers (mm)

Specks(mm)

Masses (mm)

ACR 156 FFDM ACR 156 FFDM ACR 156 FFDM

1.56

1.12 0.54 2.00

0.89 0.89 0.40 1.00 1.00

0.75 0.75 0.32 0.33 0.75 0.75

0.61 0.28 0.50 0.50

0.54 0.54 0.24 0.23 0.38

0.40 0.40 0.20 0.25 0.25

0.30 0.16 0.17 0.20

0.14

Pass/Fail Criteria

Pass

Fail

4 32 23

3

The ACR DM Phantom Prototype

4/28/2015

17

Serial Number

*Note: Gray dot in lower left corner of wax insert is an artifact due to a bubble in wax insert.

Image of Entire Phantom

Wax Insert

Serial Number

4/28/2015

18

Expanded View of Wax Insert

Pass Criteria: 2 Fibers, 3 Specks, 2 Masses

Equivalent to SFM Phantom: 4 Fibers, 3 Specks, 3 Masses

Serial NumberPass

4/28/2015

19

•New FFDM phantom equalizes attenuation inside and outside wax insert.

•This permits evaluation of artifacts over entire phantom area with same

WW and WL used to score test objects.

Effects of Thickness Equalization

Serial Number

Benefits of Prototype Phantom Design

Provides view of entire detector – artifact evaluation

W/L optimized for test objects optimizes for artifact evaluation

Finer gradations and smaller sizes of test objects

AGD measurement & limit same as SFM – meets MQSA

Provides single image/exposure for evaluation(s)

Minimal training

Provides basis for monitor and printer QC

ACR Physics Reviewers– Can see scores and artifacts on single submitted film (or image)

– Do not need different WW/WL settings

4/28/2015

20

TechnologistSection

4/28/2015

21

4/28/2015

22

4/28/2015

23

4/28/2015

24

4/28/2015

25

4/28/2015

26

4/28/2015

27

Also:

Tally sheet

Daily Counting Sheet

4/28/2015

28

Technologist ManagementForms

4/28/2015

29

Facility Room ID

MAP ID-Unit# (00000-00) Unit Mfr & Model

AW & RW Monitor QC

Procedure: Display the AAPM TG 18-QC or SMPTE test pattern.

(Another pattern that allows relevant measurements may also be used.)

Check the monitor screens to verify they are clean.

Verify that 0%-5% and 95%-100% boxes are visible (red circles).

Verify that line-pair images are sharp and distinguishable (red rectangles).

Score the ACR DM Phantom & evaluate for artifacts (RW only).

If applicable, perform monitor calibration tests per manufacturer's recommendations.

Monitor manufacturer instructions for finding & viewing TG 18-QC test pattern:

SMPTE Test Pattern TG 18-QC Test Pattern

ACR Technique and Procedure Summaries (continued)

Facility Room ID

MAP ID-Unit# (00000-00) Unit Mfr & Model

Film Printer QC

Procedure: Required equipment - ACR DM Phantom image and densitometer

Print image acquired in "ACR DM Phantom Image Quality" test.

Print from workstation/computer/PACS used for printing most clinical films.

Do not adjust window and level settings prior to printing.

Evaluate the phantom for artifacts & score.Note: If possible, print phantom image from same x-ray unit each time for this test.

Note: Use film size most commonly used for mammography.

Note: For each printer, only print a single image from a single x-ray unit or workstation.

Date:

Dmax Note: Dmax to be measured on perimeter of film.

If not available, print a clinical image and measure

in the non-breast area.

Background OD Cavity OD

ACR Technique and Procedure Summaries (continued)

Fil

m P

rin

ter

Set

up

ID of x-ray unit or workstation for printed phantom:

Film Size (8 x 10 or 10 x 12):

Dmax

SN-1234

4/28/2015

30

4/28/2015

31

4/28/2015

32

Medical PhysicistSection

4/28/2015

33

2. ACR DM Phantom Image Quality

Facility Name Room ID, Mfr & Model

ACR DM Phantom Mfr and S/N MAP ID-Unit# (00000-00)SID (cm) Survey Date

Equipment: ACR DM Phantom (required) Phantom Setup: AEC mode:

Follow procedure in the Technologist's ACR Technique & Procedure Summaries: Paddle size (IR size):

• Use clinical technique for typical screening exam of 4.2 cm 50/50 breast Paddle type (reg or flex):

• Largest IR & paddle, 5 daN or 12 lbs, Score on AW View or selected image:

• Adjust W/L to optimize test objects, zoom & pan entire image Compression force:

• For Config 2 & 3 using kVp & mAs closest to phantom techniques AEC cell position (if avail):

Phantom patient name: Target/filter (if app): kVp (if app):

Phantom patient ID: Density setting (if app):

Image sent to which PACS? Mag factor (mag mode only):

Image receptor size

Analysis

Required: ACR DM Phantom image must be free of clinically significant artifacts.

Fiber score must be ≥2.0;; speck group score must be ≥3.0;; mass score must be ≥2.0.

SNR must be ≥40.0;; CNR must be ≥2.0;; measured wax insert distance must be 70.0 ± 14.0 mm.

Timeframe: Failures of required items must be corrected before clinical use.

Initiated (or updated) technologist's ACR Technique and Procedure Summaries form

SNR ≥40.0 (P/F)

Unit-indicated AGD (mGy)

#DIV/0!

Std dev of background

Calculated CNR

#DIV/0!

Mean background signal

Mean cavity signal

Imaging Facility Room 1, Unit Mfr, Unit Model

December 1, 2015

12345-xxPhantom Mfr 1234

70

Contact Mode

Manual - Target/Filter

Config 2


Config 3

Phantom Setup

Target/filter


Config 3

Clinical - ACR DM Phantom

Mag Mode

Clinical - ACR DM Phantom


Config 2

largest

Phantom P/F

Mass score

Speck group score

Action Limits

Overall Pass/Fail

Meas = 70.0 ±14.0 mm (P/F)

Calculated SNR

CNR ≥2.0 (P/F)

SN

R &

CN

R

R

aw I

mag

e

Fiber score

Artifacts P/F

mAs

DC offset (if applicable)

AC

R D

M P

han

tom

E

valu

atio

nR

esu

ltin

g

Tec

hn

iqu

es(if

ava

ilabl

e)

kVp

≥ ¾ border

4 - 6 specks

≥8 mm long

Full Point

≥ ½ & < ¾border

2 - 3 specks

≥5 & <8 mm

Half Point

Parallel to A-C axis (mm)Distance Measurement

Fibers

Specks

Masses

Distance

4/28/2015

34

Serial Number

Mean = 542.3 Mean = 498.5

St. Dev = 7.8

CNR

4/28/2015

35

4/28/2015

36

D = KgcsD = Mean Glandular Dose

K = Entrance surface air kerma

g = glandularity of 50%

c = corrects for difference in composition (age dependent)

s = X-ray spectrum correction (Target/Filter)

Note: g and c depend on thickness, glandularity, and HVL.

Primary Ref: D.R. Dance, et al. Additional for the Estimation of Mean Glandualar Breast Dose Using the UK Mammography Dosimetry Protocol. Physics in Medicine and Biology 45, 3225-3240, 2000.

4/28/2015

37

4/28/2015

38

4/28/2015

39

4/28/2015

40

4/28/2015

41

4/28/2015

42

4/28/2015

43

4/28/2015

44

4/28/2015

45

4/28/2015

46

ACR DM QC Manual – Status

In September 2014, draft sent to manufacturers, select technologists and medical physicists for review– We hope manufacturers will adopt this manual

Extensive comments received in October– Reviewed and incorporated as appropriate

Updated draft sent to FDA in February 2015

Comments and instructions for implementation received from FDA late April 2015

ACR DM QC Manual – Status (continued)

ACR will respond to comments

ACR must apply for FDA alternative standard under current regulations– Alternative standard will allow facilities to use this instead

of the manufacturer’s manuals

– Tomosynthesis QC to be added as Appendix after FDA approval

4/28/2015

47

Top Strengths & Take Home Points

• QC Manual is comprehensive

• Phantom is of major importance

• Most failures are artifacts

• Includes most “legacy” tests

• Includes most “current” manufacturer tests

• Accommodates mfr specific test (where app)

• Accommodates growth of QC program

• Manual can realistically be implemented

End of Presentation

Questions?

Documents

Update on ACR Digital Mammography QC Manual - Priscilla Butler