Topic 10: Gamma Camera

Question 1

What is the purpose of a gamma camera?

Answer 1

•Obtains a 2D image from a 3D distribution of radioactivity

Question 2

What is the purpose of a collimator?

Answer 2

• X / g rays emitted in all directions • Need to be able to determine origin of photons

Question 3

What types of designs of collimators can you have?

Answer 3

Parallel hole (most common) – Pin-hole – Converging – Fan-beam

Question 4

What is the downside of having a collimator?

Answer 4

Collimator reduces sensitivity of detector system

Question 5

How does changing s help us deal with different energy radiation?

Answer 5

Changing s helps us deal with different energy radiation.

Increasing s - > more lead - > reduced sensitivity (assuming all else is the same)

Question 6

Using a lower energy collimator than required does what?

Answer 6

there is inadequate lead to stop oblique incoming radiation which leads to ‘septal penetration’.

With septal penetration, oblique radiation contributes to the final imaging causing blurring and streak artefacts.

Question 7

Using a higher energy collimator than required does what?

Answer 7

• The increase in lead in higher energy collimators reduces the sensitivity of the system.
• To minimise the reduction in sensitivity, hole width is increased, which will also lead to a degradation in spatial resolution of the image.

Question 8

How do you calculate the parallel hole collimator spatial resolution?

Answer 8

Question 9

Why is it important for any collimator to get the detector as close as possible to the patient?

Answer 9

to achieve the best possible spatial resolution

Question 10

How do you improve the parallel hole collimator sensitivity?

Answer 10

Better spatial resolution is achievable with narrow and long holes.

Sensitivity (of system to incoming radiation) is improved with shorter and wider holes.

Question 11

Low energy collimator types

Answer 11

Low energy general purpose - better sensitivity

Low Energy High Resolution - have a design bias to better spatial resolution

Question 12

Choice of parallel hole collimator depends on what?

Answer 12

Septal Thickness

– Photon Energy

• Hole Length and Width

– Sensitivity vs Resolution

– Activity in Patient

– Position of the Source

– Duration of Acquisition

– Type of scan

• Static or Dynamic

– Required Resolution

Question 13

what is the effect of reducing noise by reducing the counts?

Answer 13

Reduced sensitivity or less acquisition time = lower counts

Result is more noisy image

Poorer visual contrast

Noise is due to random nature of radioactive decay (poisson noise)

Question 14

Types of non-parallel hole collimators?

Answer 14

converging collimators and

diverging collimators

Pinhole collimators

Question 15

Explain converging collimators

Answer 15

• allows small objects to use more of the field of view
• Good combination of spatial resolution and sensitivity
• A variant of this collimator is called the fanbeam collimator and is commonly used for brain imaging.

Question 16

Explain diverging collimators

Answer 16

Good for imaging objects bigger than the detector

Loss of spatial resolution and sensitivity

Sometimes used on older smaller sized detectors

Question 17

Explain pinhole collimators

Answer 17

Has one ‘pin hole’

Very good spatial resolution

Image is inverted

Huge loss in sensitivity

Good for imaging small objects (such as the thyroid or small bones)

Question 18

components of a Gamma camera detector

Answer 18

Single 9.5mm thick NaI (Tl) crystal – Surrounded by reflective material to maximise light output – Hermetically sealed to protect from moisture

• Optical window – Diffuse light emitted by crystal before it falls on PMT array – Channels light away from gaps in PMTs

• ~ 60 circular PMTs in a close packed arrangement

• Processing electronics

• Shielding – to protect from radiation entering from side/back of detector

Question 19

How do you calculate the position based on the signals

Answer 19

( sum of products of the output x distance )/ (sum of all the outputs)

in x and y direction. analogous to centre of mass calculation.

The energy of incident photon= sum of outputs.

Question 20

Without a collimator what is the intrinsic position calculation?

Answer 20

~3mm FWHM

Question 21

Spatial resolution with a collimator?

Answer 21

Question 22

So… some photons are scattered and still get through the collimator because it is still parallel!!! wahh. so how the fuck do we reduce scatter by removing those sherlock?.

Answer 22

because they fucking take a retardely slow meandering path out compared to ya primary photons, they can be noticed by the fact that they have lower energy. so its like an energy filtaaaaa.

i.e.

“because scattered energy loses energy in the scattering process ENERGY DISCRIMINATION using PULSE-HEIGHT analysis can reduce number of scattered events by using an energy window. “

Question 23

What is the issue with the energy window?

Answer 23

Some small angle scatter in energy window.

Question 24

So how do we get rid of small angle scatter thats within the energy window?

Answer 24

Dual energy window scatter correction.

Use counts in second window (W1) and approximate scattered counts as area of triangle

Subtract number of scattered counts

Triple energy window scatter correction

• Use three windows, E, W1 and W2.

Scatter = area of polygon, subtract.

Question 25

What are other uses of multiple energy windows?

Answer 25

Multiple energy windows:Some radionuclides used in nuclear medicine have more than one energy of gamma emission.

For these isotopes we can set more than one energy window to maximise the collection of information and therefore our sensitivity.

Dual isotoping:

Dual isotoping can use two separate radionuclides labelled to two different pharmaceuticals,

this allows us to image two different physiological processes simultaneously at the same time.

Question 26

Imperfections of response

Answer 26

Most non-uniformities from crystal/PMT combination - mostly PMTs.

1. PMTs although matched as well as possible will have different responses.
2. Temporal drifts in PMT output, from temperature, magnetic fields and ageing.
3. Also response changes across individual PMTs. greatest at centre of the tube, smallest at edges
4. Dead space between tubes - compounds with edge effects on PMT.

Question 27

Detector issues?

Answer 27

Pulse height analysis would be different because of drift of energy peaks, and non uniformity of response and non-linear position response.

Question 28

What do we correct for detector issues?

Answer 28

1. Stabilising PMT output - tuning
2. Energy
3. Linearity
4. Uniformity

Question 29

So how do we stabilise the PMT output - tuning?

Answer 29

• problem:Temporal drifts in PMT output - temperature, magnetic fields,* ageing**.
• Irradiate detector with uniform flux of photons e.g. from Tc-99m
• look at number of counts in two energy windows (see fig) and the counts should be the same. (you put them at high energies because at low energies scatter is more of an issue.)
• Does not ensure consistent output across PMTs - tube performance matching, other correction.

Question 30

Why is the energy response spatially dependent?

Answer 30

We lose light between tubes. and we have a really good response in the middle of the tubes.

Question 31

Energy correction across the field of view?

Answer 31

Irradiate detector (without collimator) with uniform flux of photons.

Set two narrow energy windows - one above true photopeak energy

• one below true photopeak energy

Assuming photopeak is symmetrical about centre

• counts in each window should be the same
• Alter (on pixel by pixel) basis if not

Question 32

what are four causes of non-linearity in the detector?

Answer 32

Detector response varies spatially

• different sensitivity of PMTs
• different sensitivity across PMTs and across gaps
• Electronic malfunctions
• Lightguide non-uniformities

Result: Spatial non-linearity of system

Question 33

Linearity correction

Answer 33

• Two correction matrices - pixel by pixel corrections (analogous to energy correction)
• With no collimator, apply a lead mask with slits in x, and later y directions.
• Look at deviations from lines and correct for these deviations
• Unlike energy correction, this correction is relatively independent of photon energy.
• We put in a shift. - see image below.

Question 34

Uniformity correction

Answer 34

1. Present detector with a uniform flux of photons
2. collect many >10,000 counts per pixel (i.e. low noise)
3. invert image to create correction for remaining non-uniformities
4. can be acquired with collimator (to correct for collimator non-uniformities) or without a collimator.