CT and radiography Flashcards
CT matrix size
512x512, 2 bytes/pixel, 0.5 MB
CT spatial resolution
100 um FWHM
-depends on FOV, detector size, slice thickness
daignostic CT effective dose
15 mSv
dose for CT taken for CT/PET for anatomy and attenuation correction purposes
2 mSv
what must be done to mAs if kV is increased by 15% to maintain Kair at image receptor
halve mAs
k edge of iodine
33 keV
what are filters made from
mostly Al, some from Cu (unless mammo)
lung Z and bone Z
Z = 7.5 lung
Z= 12 bone
what is beam intensity proportional to?
kV^2
mAs
CT beam quality
HVL is 6-10 mm Al
radiography beam quality
HVL is 5 mm Al
Kair at image receptor
a few uGy
Mo kedge
20 keV
CsI kedge
36 and 33 keV
Rh kedge
23 keV
Ag k edge
25 keV
W k edge
70 keV
Pb k edge
88 keV
Ba k edge
36 keV
O2 k edge
0.5 keV
Ca k edge
4 keV
film with steeper gradient means what contrast?
better contrast
what % of xrays does film absorb vs screen film?
-film absorbs 1 %
-screen film absorbs 50%
what has BEST resolution?
photoconductor
-because charge is not dispersed by detection
why is there less scatter for lower tube kV and high Z?
PE will dominate over Compton
when to use high vs low latitude
latitude= dynamic range
high latitude- chest radiography (don’t need high contrast)
dynamic range of digital detectors vs screen film
digital handles 100X higher than screen-film
max film OD
2 (1% transmittance)
Useful range is 0.3 (50% transmittance) to 2
how many bits in a byte
8
radiography and film matrix size
2560x2048, 2 bytes/pixel, 10 MB
grid ratio
strip height H/gap width D
bucky factor
dose with grid/dose without grid
OD
log(Io/I)
target film OD
1.5
base+ fog film OD
0.2
what can human eye resolve?
5 lp/mm
Nyquist frequency
limiting resolution for sampling of 2 pixels/mm is 1 lp/mm
i.e. half the sampling is limiting
CT resolution in lp/mm
0.7 lp/mm
radiography resolution
5 lp/mm
what is MTF
resolution as a function of frequency
-output to input modulation
does motion blur depend on magnification?
no but magnification increases time because focal spot is smaller, so motion can increase
relate sampling frequency to pixel size
sampling frequency = 1/pixel size
number of photons /mm2 for radiography
10^5/mm2
Rose Model
SNR>5 = lesion detected. CNR has no absolute meaning, it is only relative
SNR= contrast*rootN
N= number of photons/area
contrast=(phi-phiROI)/phi
specificity
TN/(TN+FP)
sensitivity
TP/(TP+FN)
negative predictive value
TN/(TN+FN)
positive predictive value
TP/(TP+FP)
PP= 3 ps
effect on contrast with kV
increasing kV reduces contrast but decreases mottle due to more penetration
total effect on CNR depends on Z
relate FWHM to lp/mm
lp/mm = 1/2FWHM
what do you plot on ROC curve
sensitivity vs 1- specificity
accuracy of ROC curve
(TP+TN)/(TP+TN+FP+FN)
Median kerma area product in radiography
1 Gy cm2
usual radiography skin dose
<10 mGy
entrance Kair
0.1-10 mGy
integral dose for 70 kg patient who gets 1 Gy
1 J/kg * 70 kg = 70 J
LET of xray gamma ray, beta
1 keV/um
LET of alpha particles
100 keV/um
effective dose of most radiography exams
0.1-1 mSv
effective dose of CT exams
1 mSv to over 10 mSv
scatter dose at 1 m from patient
0.1% of entrance skin dose
how to reduce heel effect
-increase anode angle
-increase SID
-decrease FS
exposure index
measure of Kair at image receptor
1 uGy is EI of 100
entrance Kair order
a few mGy
resolution of digital detectors vs screen film
3 lp/mm vs 6 lp/mm
equation for magnification
SID/SOD
HVL for kV os 60,80,100,120
2,3,4,5 mm Al
deviation index
quantifies how closely Kair at receptor matches target value
+/- 3 means exposure is double or half of target Kair
SID
usually 100 cm
180 cm for chest xray
does contrast increase with mAs?
no, but CNR increases because noise is proportional to root(mAs)
dose index
peak skin dose/IRP Kair, 0.5-0.8
what is IRP
interventional reference point
15 cm closer to focal spot than system iso
IRP Kair excludes backscatter, is conservative estimate of patient skin dose
KRP in interventional radiography
200 Gycm2
effective dose interventional radiography
30 mSv
typical peak skin dose interventional radiography
2 Gy
entrance Kair interventional radiography
3 mSv/image
receptor Kair interventional radiography
3 uGy/image
where would worker get most scatter from patient?
where beam enters patient
dose from lateral projections vs AP
double
what are bowtie filters made of
Teflon
% of incident fluence absorbed by CT detectors
90%
sinogram
projection vs xray tube angle
CT rotation speed
1 rps
effective mAs
true mAs/pitch
window/level for head scan
80/40
tissue HVL for typical CT
4 cm
angular modulation
-change current as xray tube rotates around patient
-current for lat > current for AP/PA
-current in chest < current in abdomen
how many projections does CT acquire for a single projection>
1000
CT detector width
0.625 mm
typically 64 slices in each beam width
how to improve resolution in CT
rays acquired at 0 and 180 degrees are offset by 1/2 detector width
window/level for chest
1500/-500 for lung
400/50 for mediastinum
how many shades of grey with 12 bits
2^12= 4096
window/level for liver
150/60
pitch
L/NT
<1 is oversampled
axial vs helical CT
axial: table still as tube rotates
helical: table moves as tube rotates. Reduces scan time
head CTDI
60 mGy
chest CTDI
10 mGy
abdo CTDI
15 mGy
pelvis CTDI
5 mGy
what happens to mottle if you multiply mA by 4?
mottle reduced to 1/2
embryo dose from CT
-primarily from internal patient scatter
-Pb apron wouldn’t work
-0.1 mGy for chest CT
what is kq factor
converts DLP to effective dose in CT
-depends on age, size, phantom used for CTDI
-independent of kV
cause of ring artifact
bad detector
cause of multiple ring artifacts
CT not calibrated
beam hardening artifact image
motion artifact image
windmill artifact image
film speed vs light needed for development
film speed ~ 1/light needed for development
film speed ~ 1/thickness^2
film speed~ # of light photons generated
how thick is intensifying screen
200 um
conversion efficiency of scintillator
% of absorbed energy converted into light
2-20%
CR plate -what does each color do
red light- simulate and empty electron traps
-blue light- emitted and measured
white light- used to erase
what is big gamma in film?
max slope of characteristic curve
3.32 max contrast factor
change in HD curve for overprocessed film
curve shifts to left (more sensitive)
rise in toe
less contrast (slope is less steep)
change in HD curve for underprocessed film
curve shifts right (less sensitive)
less contrast
what is detector geometric efficiency proportional to?
area/(source to detector distance)^2
amplifier gain
log(Vout/Vin)
% contrast required in CT vs radiography for detections
0.3% in CT
3% in radiography
CT improves soft tissue contrast
formula to estimate CTDI for head vs body
head- 0.2 mGy/mAs
body- 0.1 mGy/mAs
what happens to contrast as latitude increases?
contrast decreases
%mAs to use for infant vs large adult
-infant: -45% mAs
-large adult: +60% mAs
how many cervical vertebrae do humans have?
7
most appropriate exam to check for kidney stone
CT
equation for DQE
DQE = (SNRout/SNRin)^2
SNRin= N/root(N)
SNRout=N(1-exp(-ux))/root(N(1-exp(ux)))
quantum sink
-limiting stage with worse SNR as it has the fewest quanta
-cannot improve SNR without improving this stage
probability of A if A occurs m(A) times in M repetitions
integral(m(A)/M
probability of A or B
P(A)+P(B)
probability of A and B
P(A)*P(B)
probability of A given B
P(A)*P(B)/P(B)
probability density function
P(x=X)=integral(p(x)dx)
expectation value <x^n>
<x^n>=integral(x^np(x)dx
<(x-xbar>^n=integral(x-xbar)^np(x)dx
what is characteristic function
FT of PDF
what is wiener spectrum
FT of auto-correlation function
approximation of exp(-ux)
=1-ux for u«<1
noise equivalent quanta NEQ
ideal NEQ = q
more noise will make quanta appear to be reduced
DQE=NEQ/true input
does amplification improve DQE?
no because it also amplifies noise
the three film efficiencies
-xrays absorbed by screen/xrays incident = 30-80%
-number of optical photons x their energy/number of photons absorbed times their energy = 5%
-optical energy that escapes screen= 50%
characteristics of high speed film system
-lower patient dose
-more mottle because less quanta are required
-reduced latitude
-higher contrast
equation showing how scatter degrades contrast
w/o scatter, contrast = (P2-P1)/P
with scatter, contrast = (P2-P1)/(P+S)
dark field imaging
makes use of the scatter
how to calculate slice thickness in helical scanning?
-measure FWHM of bead sensitivity profile as function of z position
CT uniformity test
subtract middle from periphery
-should be within 2 HU of baseline and baseline should be within 5 HU
IGRT FBCT kV and MV dose
-0.5 to 10 mSv for kV
-10-30 mSv for MV
IGRT CBCT kV and MV dose
1-35 mSv kV
30-100 mSv MV
chest radiograph imaging parameters
85-140 kVp, 5 mA, 5 ms
abdo/pelvis radiograph imaging parameters
80 kVp, 1000 mA, 50 ms
skull radiograph imaging parameters
70 kVp, 40 mAs
CT imaging parameters
120 kVp, 750 mA, 170-240 mAs
typical radiography filter
3 mm Al
CT typical filter
2 mm Al + 0.1 mm Cu
central slice theorem
taking a one-dimensional Fourier transform of a projection is equivalent to taking the two-dimensional Fourier transform and evaluating it along one direction in frequency space.
power rating of CT
100 kW
power rating of xray
25 kW
film screen resolution vs film resolution
5 lp/mm vs 100 lp/mm
CT resolution
0.8 lp/mm
kV CBCT resolution
6-9 lp/mm
pretty sure this is wrong
MV CBCT resolution
< 0.4 lp/mm
for what voltage are tubes space charge limited
<40 kV
output proportionality to kV for CT
output proportional to kV^2.6
what are pixels in fourier space
pixels are squares which are sincs in fourier space
most common detector for a direct radiography system
a-Se
For the adult, the values of k (mSv/mGy-cm) are 0.0021, 0.0059, 0.014, 0.015, and 0.015 for adult head, neck, chest, abdomen, and pelvis, respectively. Note that for children, the corresponding coefficients are higher, indicating higher effective doses per unit of DLP for children than for adults.
SI unit of air kerma
1 Gy
