Image Quality in Cardiac Angiography

1
Image Quality in Cardiac Angiography

• L 8.1

2
Educational Objectives

1. How can image quality of cardiac angiographic
   images be assessed?
2. How useful can the quality criteria be?

3
Interventional cardiology in Europe 1992-1999
112
204
75
Rotter, EHJ 2003
4
PCI in some European Countries(1994-1999)
per million
1200 2081
239 763
825 1443
242 484
267 858
800 818
Ger Fra UK Ita Nl
Spa
EHJ 2001, 2003
5
Quality of cardiac images

• background
• cardiac cine-angiographic images should allow the
  cardiologist to evaluate the anatomic (and
  sometimes functional) details which are relevant
  for clinical decision making
• variables
• technical performance of the imaging system
• patient cooperation
• angiographic technique

6
the interventional cardiologist and quality
quality its me !!
7
Quality in invasive cardiology and scientific
societies

• Scientific societies implemented guidelines to
  guarantee adequate level of quality and
  performance of invasive cardiology
• training of operators
• quantitative standards to maintain the expertise
  in coronary angiography or angioplasty
• quality-assurance programme

Pepine, J Am Coll Cardiol 199525146 Miller,
Can J Cardiol 1996124702 Cowley, Cathet
Cardiovasc Diagn 19933014 Heupler, Cathet
Cardiovasc Diagn 199330191200 Scanlon, J Am
Coll Cardiol 1999331756824
8
Quality of cardiac images and scientific societies

• the specific problem of achieving and maintaining
  high-quality standards in angiographic imaging
• responsibility of cardiac catheterization
  laboratory directors
• involves periodic cine-angiograms review
• lesion quantification (QCA, calipers)

precise criteria have never been stated for
coronary procedures
9
do we need a method for image quality assessment
in the routine practice of diagnostic (and
interventional) cardiology
?
10
Types of technical deficiencies in 308
cineangiograms (Leape, Am Heart J 2000139106-13)
N
11
Percentage of inadequate studies by different
hospitals (Leape, Am Heart J 2000139106-13)
In 12/29 hosp. 50 of studies had deficencies 6
of these are teching hosp.
12
mean fluoroscopy time, frame number and dose-area
product (DAP) in some European centers during
coronary angiography
Country DAP (Gycm2) DAP (Gycm2) FT (min) FT (min) No. of frames No. of frames
median mean median mean median mean
Greece 38.6 46.7 5.5 7.1 1620 960
Spain 27.8 39.4 6.4 9.4 903 1596
Italy 28.2 33.5 3.0 4.2 570 610
England 28.2 33.5 3.0 4.2 570 610
Ireland 33.3 37.5 3.2 4.4 580 585
Finland 39.6 52.7 4.1 4.8 417 803
Neofotistou, ER 2003
13
DIMOND 3 data
mean number of series
projections distribution
focus-detector mean distances
14
quality evaluation of angiographic images
objective methods

• based on measurement of some physical parameters
• system transfer factor K
• spatial resolution (MTF, modulation transfer
  function)
• detective quantum efficiency (DQE)
• noise
• they are rather complex and rarely applied to
  daily practice

15
quality evaluation of angiographic images
subjective methods

• test objects or phantoms
• they are able to simulate the same radiation
  conditions as the part of the body
• they describe behaviour of radiology equipment in
  specific operating condition
• evaluation of clinical images
• allow evaluation of the overall performance
  including patients collaboration and technique

16
test objects
17
quality evaluation of angiographic images
clinical images produced in different conditions

• binary classification
• pre-defined feature identification, normal vs.
  abnormal (this is typically used with test
  objects )
• correct answer must be known
• borderline visibility
• progressive judgement in terms of quality
• variable level quality (clarity of thoracic
  calcification, arrange images in order of
  preference)
• strength of agreement by different observers
  gives indications on superiority

18
lossy compression
180
150
11
19
(No Transcript)
20
improper filtering
proper filtering
21
quality evaluation of angiographic images
limitations

• set of reference images difficult to obtain
• use limited settings where perceptibility of
  abnormal feature is under experimenters control
• quality measurement is only relative
• clinical adequacy not evaluated

22
quality evaluation of angiographic images method
of quality criteria

• quality of images is assessed in comparison to
  pre-specified criteria to comply with
• effective and relevant in clinical practice
• radiographic images (Maccia, Radiat Protect Dosim
  1995 Vañò, Br J Radiol 1995, Radiat Prot Dosim
  1998 Perlmutter, Radiat Prot Dosim 1998)
• CT scan (Calzado, Radiat Prot Dosim 1998)

23
development of Quality Criteria

• 1995-1996 GISE Società Italiana di Cardiologia
  Invasiva and AIFM Associazione Italiana di Fisica
  Biomedica
• 19962003 European Concerted Action DIMOND
  Cardiology Group (Digital Imaging Measures for
  Optimizing Radiological INformation Content and
  Dose)
• contracts FI 4P-0042DG12-WSMN, FIGM-CT-2000-00061-
  DIMOND
• http//www.dimond3.org/

24
Diagnostic requirementsadapted from EUR 16260 EN

• Image criteria
• In most cases specify important anatomical
  structures that should be visible on an image to
  aid accurate diagnosis. Some of these criteria
  depend fundamentally on correct positioning and
  cooperation of the patient or good angiographic
  technique, whereas others reflect technical
  performance of the imaging system
• Important image details
• Provide quantitative information on the minimum
  sizes at which important anatomical details
  should become visible on the image. Some of these
  anatomical details may be pathological and
  therefore may not be present (ex. mitral
  insufficiency)

25

• Objectives
• to set guidelines and give methods for the
  evaluation of image quality in
• Left Ventriculography
• Left Coronary Angiography
• Right Coronary Angiography
• Angiography of Venous Graft or Arterial Free
  Graft
• Angiography of Left Mammary Artery In Situ
• Model
• European guidelines on quality criteria for
  diagnostic radiographic images (EUR 16260 EN)
  where the diagnostic requirements and image
  criteria are settled

26

• What was not intended
• to repeat what has already been included in the
  manuals of Coronary Angiography, but to give some
  guidelines about how an angiogram should appear
  provided that good equipment and a correct
  angiographic technique are used
• Warnings
• under no circumstances should an image which
  fulfils all clinical requirements but does not
  meet all image criteria ever be rejected

EUR 16260 EN
27
definition of terms

• Clinical criteria are defined as important
  anatomical features that should be visible the
  level of visualisation is as follows
• visualization characteristic features are
  detectable, but details are not fully reproduced
  (features just visible)
• reproduction details of anatomical structures
  are visible, but not necessarily clearly defined
  (details emerging)
• visually sharp reproduction anatomical details
  are clearly defined (details clear)
• Technical criteria
• help to asses the technical quality of the
  procedure
• features not necessarily impair the clinical
  information content (panning, arms position,
  etc.)
• Aspects of an optimised angiographic technique
• set of technical information
• aimed to an optimised radiological technique
• not mandatory

28
visualization characteristic features are
detectable, but details are not fully reproduced
(features just visible)
29
reproduction details of anatomical structures
are visible, but not necessarily clearly defined
(details emerging)
30
visually sharp reproduction anatomical details
are clearly defined (details clear)
31
clinical criteria for RCA projections based on
operators choice

• 1) Visually sharp reproduction of the origin,
  proximal, mid (especially the crux region) and
  distal portion in at least two orthogonal views,
  with minimal foreshortening and overlap
• 2) Visually sharp reproduction of side branches ?
  1.5 mm in at least two orthogonal views, with
  minimal foreshortening and overlap. The origin
  should be seen in at least one projection
• 3) Visually sharp reproduction of lesions in
  vessels ? 1.5 mm in at least two orthogonal
  views, with minimal foreshortening and overlap
• 4) Visualization of collateral circulation when
  present

32
technical criteria

• 1) Simultaneous and full opacification of the
  vessel lumen at least until the first
  flow-limiting lesion (in general 90-95 by
  visual estimation)
• 2) Performed at full inspiration if necessary to
  avoid diaphragm superimposition or to change
  anatomic relationship (in apnoea in any case)
• 3) Arms should be raised clear of the
  angiographic field
• 4) Panning should be limited. If necessary, pan
  in steps rather than continuously, or make
  subsequent cine runs to record remote structures
• 5) When clinical criteria 1-4 have been
  fulfilled, avoid extra projections (mainly LAO
  semi-axial)

33
aspects of an optimised angiographic technique

• 1)  Use of the wedge filter on bright peripheral
  areas
• 2) 2-3 sequences (except for difficult anatomic
  details)
• 3) 12.5-15 frames/s (25-30 only if heart rate
  exceeds 90-100 bpm or in paediatric patients)
• 4) 60 images per sequence at average (12.5-15
  fr/s) except if collaterals have to be imaged or
  in case of slow flow

34
questions on DIMOND Quality Criteria

• Are these criteria, derived from a model studied
  for static radiological imaging, suitable for the
  more complex cine-angiogram examinations ?
• Based on these criteria, is it possible to
  evaluate and quantify quality in an objective way
  ?

35
problems related to subjective evaluation of
images
36
problems related to subjective evaluation of
images
37
the method of image quality evaluation based on
DIMOND Quality Criteria
38
the method of image quality evaluation based on
DIMOND Quality Criteria
39
the method of image quality evaluation based on
DIMOND Quality Criteria
40
example of quality score calculation (QS) for RCA
41
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
42
example of QS calculation for RCA
?
?
?
?
?
?
?
?
?
sum of scores 91 (actual score) maximum
theoretical score 96 QS actual
score/theoretical score 65/88x100 94
?
?
?
?
?
?
?
43
total score (mean and std dev.)15 angio, 65
readings, 3 european centers
within pts variability 0.08 Lins coeff .76
(CI .67-.84)

AJC, 1999 (abs)
44
total score (mean and std dev.)30 angio, 160
readings, 6 european centers

45
total scores compared to subjective opinion
good and acceptable
two cases lacking
46
what is ?

• good
• I get all the information needed to treat the
  patient and I like this examination
• acceptable
• I get all the information needed to treat the
  patient but I dont like very much this
  examination
• unacceptable
• I dont get all the information needed to treat
  the patient and I dont like this examination at
  all

47
Remarks

• the method based on Quality Criteria applies to
  cardiac angiography
• reproducibility is good
• measure of clinical acceptability seems improved
  in comparison to subjective opinion
• the method forces to a systematic and
  standardized analysis of the images
• specific training not requested (but it may
  improve agreement)

48
Quality Criteria published papers

• Criteri di Qualità dellImmagine Cineangiografica
  (documento preliminare). Emodinamica 1997 10
  (suppl.) 9-11
• Quality criteria of imaging in diagnostic and
  interventional cardiology. TCT-196 Am J Cardiol,
  199984(6A)73P-74P
• A method based on DIMOND Quality Criteria to
  evaluate imaging in diagnostic and interventional
  cardiology. Radiat Prot Dosim 200194167-172
• Quality Criteria for cardiac images in diagnostic
  and interventional cardiology. Br J Radiol 2001
  74852-855

49
closing remarks

• image quality is not warranted in coronary
  angiography
• a great variability is found in common practice
  among different operators and radiological
  exposure varies considerably
• image quality assessment plays a pivotal role in
  the optimisation of angiographic procedures
• optimisation implies a continuous process of
  research and audit which should involve
• Scientific Societies
• single operators
• cooperation of all professionals in the Cath. Lab.