MAMMOGRAPHY

- HIGH LEVEL
TROUBLESHOOTING

Maynard High
New York Medical College

SS2001-M.High 1
Objectives:
Objectives
• Review MQSA and ACR annual QC tests as
opportunities for troubleshooting before a
significant image quality problem develops.
• Review MQSA equipment evaluations, and
the role they play in preventing the need for
troubleshooting at a later date.
• Present a logical decision tree approach to
troubleshooting specific image quality
problems.
SS2001-M.High 2
 The Medical Physicist
CAN & SHOULD
be an important resource for
improving image quality of:
• Phantom images
&
• Clinical images

SS2001-M.High 3
Medical Physicist’s Services:
Compliance or Quality?
• Growing tendency toward compliance
– easier and quicker to meet minimum standards
than to improve on them
– “the MQSA inspector accepts it”
• Both ACR and MQSA encourage the
physicist to go beyond compliance checks,
and to make recommendations for quality
improvement.
SS2001-M.High 4
Medical Physicist’s Services:
Compliance or Quality?
• During annual testing, make suggestions
that may lead to improved image quality.
– this may prevent more intensive troubleshooting
in the future
• Example:
– You see the maximum compression force
is set to 30#. Although compliant, you
should advise increasing the max to 45#.
SS2001-M.High 5
Physicist Level Performance
Evaluations & Troubleshooting:
• ANNUAL quality control compliance
testing
– MQSA required testing
– ACR recommended testing
• Equipment evaluations
– New, repaired, relocated equipment
– MQSA required

SS2001-M.High 6
Physicist Level Performance
Evaluations & Troubleshooting:
• Troubleshooting (specific problems)
– Improving contrast, detail, noise, artifacts, etc
– Goes beyond performance testing of x-ray units
• screen-film appropriateness
• film processing evaluation
• technique factors
• proper use of AEC
• viewing conditions

SS2001-M.High 7
Medical Physicist’s Services in
Mammography:
• ANNUAL quality control compliance
testing

• Equipment evaluations

• Troubleshooting

SS2001-M.High 8
MQSA Required Annual
Physicist QC Survey:
• The 12 equipment tests in Table 1
– testing methods not generally specified
– see Guidance (www.fda.gov/cdrh/mammography)
• Evaluation of technologist QC testing
• A written summary report
– format of report not specified
– to include recommendations physicist feels may
improve the quality of mammography.
SS2001-M.High 9
ACR Recommended Annual
Physicist QC Tests:
• The 11 equipment tests in Table 2
– testing methods detailed in 1999 ACR
Mammography Quality Control Manual
– tests, action limits, scope of testing are very
similar to MQSA, but there are several
differences which Medical Physicist must
consider, and use profesional judgement to
develop a comprehensive performance evaluation.

SS2001-M.High 10
ACR Recommended Annual
Physicist QC Tests:
• Evaluation of technologist QC testing
• A written report including a summary
– model report formats are in the ACR manual
– to include recommendations physicist feels may
improve the quality of mammography.

SS2001-M.High 11
Resources:

SS2001-M.High 12
Medical Physicist’s Professional
Judgment: Tailoring the Tests
• Guidance documents can never set the
scope of testing and limits for all
mammography units because:
– many different existing designs
– new designs with new features
– State Department of Health requirements
• Professional judgment & knowledge of unit
must be used to tailor the tests to the unit
SS2001-M.High 13
Multiple Selectable mA Stations:

• All tests affected by mA should be

performed at each clinically used mA
station
– kVp
– radiation output rate
– system resolution
– small focus is always a different mA station

SS2001-M.High 14
Multiple Filters (Mo, Rh, Al):

• All tests affected by filtration should be

performed for each clinically used filter
– HVL
– average glandular dose
– phantom image quality
– AEC performance
– artifact evaluation

SS2001-M.High 15
 Multiple Targets (Mo, Rh, W):

• All tests affected by anode selection should be

performed for each clinically used target
– HVL
– average glandular dose
– phantom image quality
– AEC performance
– artifact evaluation

SS2001-M.High 16
Multiple Auto-kV, Auto-filter
Algorithms:
• Selectable maximum exposure time (0.8, 1.6s)
• Selectable dose (STD, CNT, DOSE)
– HVL
– average glandular dose
– phantom image quality
– AEC performance
• Test pulses may affect dosemeters
SS2001-M.High 17
Multiple AEC Sensors:

• Some units use separate sensors for small

and large Bucky’s
– Must test thickness/kVp tracking for each

SS2001-M.High 18
Multiple AEC Sensors:

• Some units use a matrix of fixed detectors

which can be used to find the densest
portion of the breast
– Must test each detector independently in some way

SS2001-M.High 19
Multiple Film-Screen Combos:

• Some facilities use a faster speed screen for

mag and/or large Bucky
– uniformity of screen speed
– average glandular dose
– phantom image quality
– AEC performance
– system resolution

SS2001-M.High 20
Medical Physicist’s Professional
Judgment: Tailoring the Limits
• Action limits can often be tightened
– kVp accuracy & reproducibility
– AEC performance can be made better than
+/- 0.15 OD for 2 - 8 cm on some units
• Action limits can be tailored at acceptance
testing.

SS2001-M.High 21
The Annual Medical Physicist’s
Survey Report: Requirements
• Communication of results with
recommendations is just as important as
performing the tests.
• Needed on-site for annual MQSA
inspections
• Needed to be submitted with applications
for accreditation.
SS2001-M.High 22
The Medical Physicist’s Report:
Useful for -
• Information resource for technologist
– Image quality and dose of Rh filter, AEC mode
• Information resource for service engineer.
– Details of calibration deficiencies
• Information resource for troubleshooting
– Performance baseline to which future
measurements can be compared

SS2001-M.High 23
The Medical Physicist’s Report:
Useful only if -
• There is sufficient information in report.
– Adequate description of test conditions
• kVp, mA, filter, target
• SID, magnification, receptor size
• phantom type and thickness
• AEC settings, compression force
• screen, film, processing conditions
– Test results for complete range of machine
settings and modes of operation clinically used
SS2001-M.High 24
The Medical Physicist’s Report:
Useful only if -
• Contains detailed recommendations for
corrective action.
• And contains:
“ …recommendations that will improve
image quality, including [those] concerning
image receptors, technique factors,
processing, viewing conditions and
technologist QC.” (ACR QC Manual)
SS2001-M.High 25
Medical Physicist’s Services in
Mammography:
• ANNUAL quality control compliance testing

• Equipment evaluations

• Troubleshooting

SS2001-M.High 26
Equipment Evaluations:
Evaluations
Requirements
• Before use on patients, a qualified medical
physicist must perform an MQSA
“equipment evaluation” on :
– newly installed mammography equipment (x-ray
unit or processor)
– disassembled and reassembled equipment
– equipment which has had a major component
changed or repaired.
SS2001-M.High 27
Equipment Evaluations:
New Equipment Acceptance Test
• ACR Mammography QC Manual:
“It is assumed that mammography equipment
will have been subjected to more extensive
acceptance testing or a thorough performance
evaluation prior to initiation of [annual] QC
testing.”

SS2001-M.High 28
Equipment Evaluations:
New Equipment Acceptance Test
• MQSA Compliance Guidance
“The equipment evaluation is more extensive
than the survey. It may be regarded as an
acceptance test for equipment and an annual
survey alone is not sufficient to meet this
requirement.”

SS2001-M.High 29
Equipment Evaluations:
Disassembled, Reassembled
• Relocated, disassembled and reassembled
equipment is to be evaluated as newly
installed.
• A complete acceptance test needed

SS2001-M.High 30
Equipment Evaluations:
“Major” Repairs
• New x-ray tube
• AEC component replacement
• Collimator replacement
• Beam filter replacement
• Processor reassembly
• Generator replacement or re-calibration
are considered major repairs.
SS2001-M.High 31
Equipment Evaluations:
Scope for “Major” Repairs

• Tests would include all those affected by

the component repaired or replaced.

SS2001-M.High 32
Equipment Evaluations:
Example - New AEC sensor
• AEC testing
• Dose determination
• Phantom tests

SS2001-M.High 33
Equipment Evaluations:
Rebuilt/Replaced Processor
• Sensitometric testing
• Phantom tests
• Artifact evaluation
• Dose determination
• AEC testing
• Verification of proper processing solutions

SS2001-M.High 34
Equipment Repairs:
Physicist Involvement
• Medical physicist evaluates in person
• Medical physicist provides oversight
• Medical physicist involvement optional

• Should be discussion between facility and

medical physicist

SS2001-M.High 35
 The Mammography Quality
Standards Act Final Regulations
Document #4
May 23, 2001

www.fda.gov/cdrh/mammography

SS2001-M.High 36
 From CDRH Guidance Document #4
Table: Medical Physicist Involvement in Equipment Adjustments,
Changes, or Repairs
For any adjustment, change, or repair not listed in the table below, or if the facility is unsure as to the full
extent of the adjustment, change, or repair, the facility should consult their medical physicist to determine
the proper extent of his or her involvement in evaluating the item.
Medical Physicist
Item Major Repair
Involvement
Automatic Exposure Control
MP conducts evaluation
AEC Replacement Y
in person
Thickness compensation internal* MP conducts evaluation
Y
adjustment in person
MP conducts evaluation
AEC sensor replacement Y
in person
MP conducts evaluation
AEC circuit board replacement Y
in person
Density control - internal* adjustment N MP oversight
Bucky (New to Facility) Replacement
MP conducts evaluation
AEC also replaced Y
in person
AEC not replaced N MP oversight
SS2001-M.High 37
Medical Physicist’s Services in
Mammography:
• ANNUAL quality control compliance testing

• Equipment evaluations

• Troubleshooting

SS2001-M.High 38
Manufacturer’s Troubleshooting
Resources:
• Film processor service company
• Film manufacturer representative
• Mammography unit service engineer
• Mammography unit applications person

• Should be advised where appropriate

SS2001-M.High 39
Manufacturer’s Troubleshooting
Resources may not resolve issue -
• Poorly trained service person
• Narrow focus and expertise of individual
• Management decisions resulting in poorly
compatible system components
– film/screen/chemistry/processor
• Physicist can analyze entire imaging
system to determine where improvement
can
SS2001-M.High
be made. 40
Troubleshooting by the Medical
Physicist
• Involvement of medical physicist in solving
specific problems concerning image quality
• Will evaluate entire imaging chain
– Film processing
– Film-Screen combination
– X-ray unit
– Technologist technique
SS2001-M.High 41
Triggering Events for Physicist’s
Troubleshooting: #1
• Failure of Technologist’s QC test
– processor sensitometry
– phantom image quality

• Physicist must be knowledgeable in

performing technologist’s tests and in all
variables which affect test results.
SS2001-M.High 42
Triggering Events for Physicist’s
Troubleshooting: #2
• Patient’s current mammograms perceived
to be not as good as previous images.
• Physicist must
– separate reality from impression
– establish which image quality parameter degraded
– establish possible cause of degraded parameter
– make recommendations for improvement

SS2001-M.High 43
Triggering Events for Physicist’s
Troubleshooting: #3
• Facility fails its application for
accreditation because of image quality

• This risk can be reduced if physicist

reviews all images before submission.
• Of course physicist must be familiar with
the image quality criteria used by the ACR
for both phantom and clinical images
SS2001-M.High 44
ACR Image Quality Criteria for
Clinical Images:
• ACR Mammography QC Manual
• Accreditation Application Package
• Accreditation Results Report returned to
facility
• Literature (Bassett et al. 2000)

SS2001-M.High 45
Clinical Image Failures : 1997*
• Positioning 20%
• Exposure 15%
• Compression 14%
• Sharpness 13%
• Contrast 13%
• Artifacts 11%
• Labeling 8%
• Noise 5%
SS2001-M.High *Bassett, et al;Radiology 46
2000;215:698-702
 Inadequate
pectoralis on

22%
MLO

SS2001-M.High
Sagging
breast on
14%
MLO

Poor
visualization
post. tissue
14%

on MLO

698-702
Skin folds
12%

*Bassett, et al; Radiology 2000; 215:

Poor
visualization
post. tissue
12%

on CC

Post. Nipple
20 %

line on CC
10%
Positioning Problems : 1997*

<1cm of MLO
47
Posterior nipple line on CC not
within 1 cm of that on MLO

6.5 cm

0 cm
8.

SS2001-M.High 48
Mammography Troubleshooting
Categories:

• Film processing
• Film-Screen combination
• X-ray unit
• Technologist technique

SS2001-M.High 49
Troubleshooting the Film
Processing: General principles
• Film must be processed as recommended
by film manufacturer (or equivalently)
– physicist can compare processing at another
site using recommended processing.
• Sensitometry does not guarantee proper
processing, only consistency.

SS2001-M.High 50
Troubleshooting the Film
Processing: General principles
• Look for changes
– film type, emulsion number
– chemistry type, supplier, mixer
– staffing, hours of operation, films/day
– PM schedule, PM personnel
• Film processing must be evaluated before
troubleshooting the x-ray unit
SS2001-M.High 51
Troubleshooting Resources:

SS2001-M.High 52
Troubleshooting Resources:

SS2001-M.High 53
Troubleshooting Resources:

SS2001-M.High 54
Mammography Troubleshooting
Categories:

• Film processing
• Film-Screen combination
• X-ray unit
• Technologist technique

SS2001-M.High 55
Troubleshooting the Film-Screen
Combination: High system speed
• Analysis of 31,000 phantom images
submitted to ACR showed:
High failure for masses & fibers (30-40%)
associated with doses lower than 0.75 mGy
(Haus, Yaffe, Feig, et al. 2000)
• Low doses should trigger a careful
evaluation of image noise.

SS2001-M.High 56
Troubleshooting the Film-Screen
Combination: Low System Speed
• May result in blur due to long exposure
times
• May result in low contrast if kVp is raised
to achieve proper exposure times
• Dose and risk may be
needlessly high.

SS2001-M.High 57
Average Glandular Dose:

• An important troubleshooting tool

• It should be within an appropriate range

SS2001-M.High 58
Mammography Troubleshooting
Categories:

• Film processing
• Film-Screen combination
• X-ray unit
• Technologist technique

SS2001-M.High 59
Troubleshooting the
Mammographic X-Ray Unit:
• Physicist may feel this is an easy category
– Make a few measurements
– Compare with last report
– Voila! The offending parameter pops out.

• Things are seldom so simple

SS2001-M.High 60
Example: Troubleshooting AEC
because of Low OD on Images
• Seldom fruitful to have service change
AEC calibration without determining
reason for low OD

• Need a troubleshooting algorithm

SS2001-M.High 61
Troubleshooting Low OD:
A possible algorithm
First, get the facts:
• Under what conditions were images light?
– All images, parts of images, dense, fatty, thick,
thin, high kVp, Rh filter, large Bucky only, etc
• Was there any correlation with
– x-ray unit, cassette #, technologist, radiologist,
time of day, day of week, PM schedule, etc

SS2001-M.High 62
 Areas of film with OD < 1.0 are
UNDER-EXPOSED

OD = 1.6

OD = 0.4

SS2001-M.High 63
Troubleshooting Low OD:
A possible algorithm
Next, check film and film processing:
• Nothing can be learned about AEC if
processing is not in control
• Any film, chemistry, processing changes?
• Review sensitometry records
• Is processing in control today so AEC
testing will be meaningful?
SS2001-M.High 64
Troubleshooting Low OD:
A possible algorithm
Next, check phantom image QC records:
• Has mAs been stable indicating stability of
AEC?
• Review phantom OD plot for stability
• If there are changes, are they correlated
with changes in film sensitometry?
• If needed, go back to original images.
SS2001-M.High 65
Troubleshooting Low OD:
A possible algorithm
Next, review technologist technique:
• Check Technique chart
• Query all technologists about AEC modes
used, filters chosen, sensor positioning
• Compression adequate to spread tissue?
• Physicist needs working knowledge of the
various modes available on machine.
SS2001-M.High 66
Troubleshooting Low OD:
A possible algorithm
Review inputs used by AEC:
– Thickness (position of compression paddle)
– Compression force
– kVp
– mA, exposure time
– Attenuation measurements (test pulse)
– Film type selected
– Dose or time mode chosen
SS2001-M.High 67
Troubleshooting Low OD:
A possible algorithm
Now expose some phantoms:
• Simulate patient exposure by using normal
compression force with normally used
AEC mode.
• Process films as technologists do
– emulsion down
– lengthwise on right side of feed tray

SS2001-M.High 68
Film must be fed properly

MinR 2000 Emulsion MinR 2000 Emulsion

side down side up

OD=1.51 OD=1.41
SS2001-M.High 69
Troubleshooting Low OD:
A possible algorithm
Analyze the data:
• What is the proper OD?
– Phantom > 1.4, Patient >1.0 in densest portion
• But, Depends on
– type of film
– viewbox luminance
– radiologist
SS2001-M.High 70
Troubleshooting Low OD:
A possible algorithm
Finally review all information and make
recommendations:
• There may be more than one cause.
• Areas of possible improvement unrelated
to the light films may be uncovered.
– Make recommendations concerning these

SS2001-M.High 71
 LESSON LEARNED #1
Troubleshooting in
mammography generally
involves several problem
sources.
They all need to be considered.

SS2001-M.High 72
 LESSON LEARNED #2
The medical physicist must be
knowledgeable about
mammography equipment
design and how it is used
clinically.

SS2001-M.High 73
 LESSON LEARNED #3
The medical physicist must
develop a logical and efficient
troubleshooting algorithm.

SS2001-M.High 74
Troubleshooting Decision Trees:

• Identify the particular image quality

problem.
• Make a list of the possible causes or
contributing factors.
• Create a decision tree algorithm to test
causes.

SS2001-M.High 75
Sensitometry Density and/or
Density Difference Change:
• Film emulsion # • Improperly mixed
(use reserved film) developer.
• Developer temp • Change in developer
• Immersion time type/brand
• Replenish rates • Oxidized developer
• Contaminated dev. • Expired/improperly
stored film

SS2001-M.High 76
Sensitometry Base + Fog Increase

• Contaminated • Light leak

developer

• Expired/improperly
stored film

• Improper safelight

SS2001-M.High 77
Phantom OD or Contrast Change

• Film emulsion # • kVp calibration or

(use clinical film) setting

• Target/filter setting
• See all items for
sensitometry • Cassette/screen
changed
• AEC calibration or
setting • Position of phantom
& AEC sensor
SS2001-M.High 78
Phantom mAs Change

• AEC calibration or • Cassette/screen

setting changed

• kVp calibration or • Position of phantom

setting & AEC sensor

• Target/filter setting

SS2001-M.High 79
Phantom Score Change

• See all items for • See items for low

sensitometry contrast, blur and
noise
• See all items for
Phantom OD or
Contrast change

• Excessive artifacts
SS2001-M.High 80
Low Contrast on Patient Images

• Most common cause

generalized under-
exposure, or under- OD = 0.4
penetration of dense
portions of breast

• AEC sensor not

under densest
portion of breast.
SS2001-M.High 81
 Exposure Problems : 1997*
15%

46%
40%

9%

Generalized Inadequate Generalized

under-exposure penetration of over- exposure
dense areas

SS2001-M.High *Basset,et al;Radiology 82

2000;215:698-702
 Contrast Problems : 1997*
13%
89%

10%

Inadequate Contrast Excessive Contrast

SS2001-M.High *Bassett,et al;Radiology 83

2000;215:698-702
Low Contrast on Patient Images

• See all items for • Film processing does

sensitometry not meet
manufacturer’s
• See all items for recommendations
Phantom OD or
Contrast change

• High fog level

SS2001-M.High 84
 Low Contrast Case Study #1
Chemistry changed to non film
manufacturer by department
manager, but not caught by
sensitometry or phantom tests
because mistakes were made in
plotting.
Non-optimal film processing.
SS2001-M.High 85
Low Contrast on Patient Images

• kVp calibration or • AEC calibration or

setting (may be setting leading to low
affected by AEC OD (may affect kVp
mode) or target/filter)

• Target/filter setting • Inadequate

(may be affected by compression
AEC mode)

SS2001-M.High 86
Compression Force is often
operator adjustable

SS2001-M.High 87
Deflection Should be < 1 cm

6.5 cm 5.5 cm

SS2001-M.High 88
 Low Contrast Case Study #2
Compression force
miscalibrated, indicated 20 dN,
measured <25 #
Resulted in 9 cm compressed
thickness (compared with 5 cm
previous year) and AOP
changed filter to Rh.
SS2001-M.High 89
Excessive Blur on Patient Images

• Marginal focal spot • Long exposure times

performance (kVp may be too low)
time should be
• Inadequate between 1 - 2 sec.
compression (at least
25#, but as much as • Poor film-screen
possible) contact

SS2001-M.High 90
 Compression Problems : 1997*
59% 14%

23%

9%

Poor separation of Patient motion Non uniform

parenchymal exposure levels
tissues

SS2001-M.High *Bassett, et al; Radiology 2000; 215: 91

698-702
Excessive Noise on Patient
Images
• Film-screen • Film processing does
combination too fast not meet
manufacturer’s
• Correlates with low recommendations
Average Glandular (this is a special
Dose concern in case of
non-dedicated
processor)
SS2001-M.High 92
 Noise Problems : 1997*
5%

74%

26%

Visually Striking Mottle Limited Visualization of

Pattern Detail because of Noise

SS2001-M.High *Bassett, et al; Radiology 2000; 215: 93

698-702
Excessive Artifacts on Patient
Images
• X-ray unit filter • Dirty/worn or
• Screen artifacts misaligned rollers
• Film handling • Nightly cleaning of
• Darkroom dust cross-over rollers
• Improperly installed • Not following
or vented processor manufacturer’s
recommendations for
processing

SS2001-M.High 94
 Artifact Problems : 1997*
11%

37%

29%

12% 12%
9%

Dirt or Scratches Grid Roller Other

Lint or Pickoff Marks

SS2001-M.High *Bassett, et al; Radiology 2000; 215: 95

698-702
Mammography Troubleshooting
Categories:

• Film processing
• Film-Screen combination
• X-ray unit
• Technologist technique

SS2001-M.High 96
Technologist skill affects image
quality:
• Responsible for positioning & compression
• Controls OD by AEC sensor positioning
• Controls contrast by kVp and target/filter
selection
• Controls motion blur (exposure time) by
kVp selection

SS2001-M.High 97
Compression is one of most
important image quality factors
• Separates structures within breast
• Reduces thickness of breast
– more uniform OD
– less motion
– shorter exposure time
– reduced geometric blur

SS2001-M.High 98
Physicist can & should evaluate
compression
• Is compression force properly set and
calibrated?
– Physicist should perform test personally and
compare with technologist’s records

• Does compression force hold for length of

time it takes to complete patient exposure?

SS2001-M.High 99
Physicist can & should evaluate
compression
• Is compression mode used properly by
technologist?
– Some units have programmable, multi-step
compression modes, that if not understood,
can result in incomplete compression.

SS2001-M.High 100
Physicist can & should evaluate
compression
• Is the technologist using adequate force?
– Many units print compression force on the
film or on a sticker
– The physicist should review patient films to see
that at least 25# of force is being applied.

SS2001-M.High 101
kVp controls both contrast and
exposure time
RULE of THUMB

Maximize contrast by selecting the lowest

kVp consistent with an exposure time
between 1 and 2 sec to reduce motion
blurring.

SS2001-M.High 102
kVp controls both contrast and
exposure time
• Is the technologist using a kVp which
properly balances contrast and time?
• Review auto-kVp AEC modes used with
technologists and explain their influence on
kVp
• Review exposure times on films
• Consult technique charts
SS2001-M.High 103
Technique charts are valuable

• Is technique chart conspicuously posted?

• If not, images may not be consistent among
all technologists.

• Is technique chart current?

• Chart may not have been changed to
reflect changes in screen/film/chemistry
SS2001-M.High 104
Technique charts are valuable

• Do all technologists follow chart?

• Need to query all technologists.

• Are recommended settings appropriate?

• Do AEC phantom tests result in an
acceptable exposure time and kVp?

SS2001-M.High 105
Mammography Image Quality
Troubleshooting
• Requires good acceptance and periodic QC
testing data
• Requires analysis of entire imaging chain
– Film processing
– Film-Screen combination
– X-ray unit
– Technologist technique

SS2001-M.High 106
Mammography Image Quality
Troubleshooting
• The medical physicist needs to expand
his/her knowledge and expertise beyond x-
ray unit testing to be a more valuable
resource for improving image quality.

SS2001-M.High 107
My recommended solution to
film processing problems:
*

* May not be FDA approved

SS2001-M.High 108