Transcript Slide 1

Planar scintigraphy produces two-dimensional
images of three dimensional objects. It is handicapped by the
superposition of active and nonactive layers which restricts
the accurate measurement of organ functions.
Emission computed tomography (ECT) is based on the
production of multi cross sectional images of tissue function
which are used to produce by overlay three dimensional images.
Two ECT techniques are currently available:
• Single Photon Emission Tomography (SPECT)
which involves the imaging of single 7-ray activity
(typically from 99Tcm).
• Positron Emission Tomography (PET) which involves
the imaging of the 511 keV annihilation radiation
originated in positron decay (typically from 18F)
Image planes are derived by using two different techniques:
• longitudinal ECT (limited-angle technique), photons are
detected within a limited angular range from several body
sections simultaneously. The reconstructed image planes
are positioned parallel tothe detector plane.
• transaxial ECT (transverse section technique), the
detector moves by 360 ° around the body to sample
photons from multiple body sections. The reconstructed
image planes are perpendicular to the detector plane.
Longitudinal ECT allows to view the
radioactivity from different angles (within a limited angle
range) to obtain information about the depth of the
radiation source.
This is done by using a
rectilinear scanner system coupled
with a highly focused collimator who
creates a sharp image only from a
particular plane at a depth defined
by the angle range (focal point).
The multiplane tomographic
scanner represents an improved
version which replaces the single
detector by a gamma camera. It also
allows to adjust (by electronic
repositioning) the focal distance and
can therefore select different image
planes.
As example are shown twelve images of the skeleton
recorded in longitudinal multiplane tomography technique.
Each image represents a different plane along the body.
Single-Photon Emission Computed Tomography
(SPECT) with a rotating gamma camera (transaxial ECT)
allows multiple views of the three dimensional distribution of the
radioactivity from different directions.
The gamma camera is
coupled to a parallel hole collimator (no
focusing) which allows to produce a 2D
image (64x64) consisting of multiple
profiles (64), each profile represents a
ID projection of the radioactivity in the
profile (distributed over 64 channels).
Each point in the profile represents the sum of the activity
along the line of sight:
with € as detector efficiency and  as solid angle
The camera rotates either continuously or in fixed angle steps
and repeats the monitoring until the completion of the 360° turn.
The three dimensional image is constructed by using similar
Fourier analysis techniques as designed for X-ray CT scanning.
Positron Emission Tomography (PET) operates by using
at least two opposite to each other positioned rotateable detector.
PET is based on the principle of detecting annihilation
radiation with coincidence techniques.
The injected radionuclide must be a positron (+) emitter. The
positron annihilates after about 1mm path length (depending on density of
tissue material and on the energy of the positron) and emits two 511 keV
photons in opposite directions.
Detection of both photons in coincidence
defines a line along which the annihilation event has
taken place. The position of the radionuclide is within
1mm distance. This distance as well as a slight
deviation from the 1800 emission of the two photons
limits the spatial resolution to about 1mm – 2mm.
The use of annihilation radiation coincidence technique
in PET improves the quality of image formation considerably
compared to collimator techniques used in SPECT.
In SPECT the intensity and the resolution of the  signal
degrades with increasing depth, due to attenuation through body
tissue of increasing thickness d,
and due to the degradation of collimator resolution €c with
increasing source collimator distance z:
with constant hole diameter d and hole length L for the collimator
system.
In annihilation radiation coincidence measurements
the resolution remains essential constant with depth because
of the uniformity of the geometric response defined by the
straight line between the two detection processes.
The intensity of the coincidence signal is defined by the
attenuation in body material from the point of annihilation at depth d
in both directions,
with T being the thickness of the body along the line and 
(x)dx .
Therefore the intensity for the annihilation signal along the
line is independent of the depth.
The absolute count rate for coincidence events is determined
by the count rate for true coincidences Itrue (real coincidence events
originating from one single annihilation process) and for random
coincidences Irandom (fake coincidence events which occur when
accidentally each detector records an uncorrelated signal within a time
window  ).
The count rate for true coincidences from I0
annihilation events is determined by the efficiency e and
solid angle  of each detector:
for present PET machines the total efficiency for
coincidence measurement,
The random coincidence count rate is determined by the
count rate for single events:
in the two detectors and by the coincidence time window r:
This yields a ratio of true to random coincidences:
which is independent of efficiencies and solid angle but only depends
on the intensity of the emitted annihilation radiation, and the
coincidence time window:
At these conditions the intensity of the radiation source inside the body
must be at least I0 106 events/s to obtain a true to random ratio of unity.
This would require a source strength of
at least 1 MBq inside the body. A 1 MBq source
inside the body (neglecting attenuation effects
corresponds to a random count rate of:
To improve the detection conditions and to separate the true
from the random coincidences the time structure of the signals can be
utilized by decreasing .
Separation is based on the fact that random events come continuously
while true events come within a few nano seconds (speed of light c=3*1010 cm/s).
Standard electronic
is used to separate true
coincidence events in the
sharp time peak from random
events using fast timing
conditions on the electronic
signals.
For  = 10 ns at a count rate of I0 = 106 s-1 a true coincidence rate of:
is obtained at a random coincidence rate of:
Modern PET systems are based on multicrystal designs.
Instead of a rotateable detector pair
the patient is surrounded by a ring of individual
small Nal scintillator detectors. Each detector
is coupled to his own individual phototube and
is in electronic coincidence with any of those
detectors at the opposite site of the patient.
With such a device a multiple image can be obtained in one shot.
This is particular important for monitoring
physiological processes with short time scales
( t  10-3 s).
Detecting coincidence events between one detector and two
neighboring detectors at the opposite site defines the spatial
resolution of the device to  5mm.