Digital subtraction angiography

Question 1

what defects can be visualised in angiography?

Answer 1

It is possible to visualise stenosis, aneurysms and other defects in the vasculature. Angiography requires access to a high performance x-ray imaging system which can produce dynamic images depicting the flow of contrast through the vasculature of interest. Advances in angiographic imaging platforms have also made it possible to perform image-guided interventional therapies to re-establish adequate blood flow.

Question 2

HOw does DSA work?

Answer 2

It uses a specialised form of digital fluorography system to capture a sequence of high quality digital images. These images are acquired/captured:
Using high intensity pulses of x-rays of short duration
Typically at an acquisition rate between 2 and 6 frames per second
In real time, then digitised

The digital fluorography system then implements the subtraction algorithm and selectively subtracts image frames in a computer.

Question 3

What is the benefit of DSA?

Answer 3

Digital subtraction of the competing anatomical structures (anatomical noise) significantly improves the contrast sensitivity in vascular imaging.

Because of the benefits of electronic signal processing and control, high quality images are produced automatically with minimum user interaction.

Question 4

What is the effect of the background in linear subtraction and why can this be a problem?

Answer 4

A blood vessel of given diameter and containing a fixed concentration of iodine may often be superimposed upon a background anatomy of widely varying x-ray attenuation. Following linear subtraction, the displayed contrast of such a vessel will vary markedly along its length:

Where the x-ray transmission of the background anatomy is high, the vessel contrast will be high
Where the background transmission is low, the vessel contrast will be low

Consequently, radiologists find that DSA images produced by linear subtraction are difficult to interpret.

Question 5

How can difficulties in linear subractio DSA be overcome?

Answer 5

The ambiguities in vessel contrast rendition produced by linear subtraction are solved by using logarithmic subtraction.

Question 6

WHat is the noise print effect in log subtraction DSA?

Answer 6

As a result of log subtraction, the noise varies across the image depending on the x-ray transmission of the background. This produces a characteristic DSA image artefact known as the noise print effect.

Radiologists have little difficulty coming to terms with the noise print effect and it normally does not disturb the diagnosis.

Question 7

What does as DSA image processor consist of?

Answer 7

High speed analogue-to-digital converter (ADC)
Two RAM arrays (M1 and M2), each capable of holding a (1024 x 1024) image
Large capacity hard-disc with 10s of gigabytes of data storage
Relevant data processing
Digital-to-analogue converter (DAC)

Question 8

What is landmarking?

Answer 8

Complete (100%) subtraction of the background anatomy is not always the preferred method of documenting the procedure.

For example, in some circumstances the surgeon may prefer to retain a low contrast anatomical framework along with the vasculature in the subtracted image to aid planning of the surgery.

This can be achieved using a processing technique known as landmarking. Landmarking is achieved by subtracting a reduced proportion (say 80 to 90%) of the reference mask image from the contrast medium enhanced image.

Question 9

A DSA system typically includes several components.

Which of the following is/are DSA components:
A. A large diameter multi-field image intensifier tube
B. A Vidicon TV camera tube
C. A DAC which is used to digitise the input video signal
D. A specialised digital processor which subtracts image data on a pixel-by-pixel basis

Answer 9

A. Correct. The input stage of a DSA system includes a large diameter, multi-field image intensifier tube. Alternatively, in modern systems a flat panel detector could be used.

B. Incorrect. A DSA system requires a low lag, wide dynamic range and low noise TV camera such as a Plumbicon TV tube or CCD camera. A Vidicon tube is unsuitable due to its high lag.

C. Incorrect. The analogue input video signal is digitised by an analogue-to-digital converter (ADC).

D. Correct. A digital processor subtracts the mask image from the contrast medium enhanced image on a pixel-by-pixel basis.

Question 10

Digital subtraction angiography systems use logarithmic subtraction of the image data in order to:
A. Minimise the number of pixels required to form the image
B. Compensate for exponential attenuation processes
C. Compensate for logarithmic attenuation processes
D. Minimise contrast loss due to x-ray scatter

Answer 10

A. Incorrect. Logarithmic subtraction has no effect on the number of pixels in the image.

B. Correct. Logarithmic subtraction compensates for the exponential attenuation processes and thus produces an image that represents the vessel contrast.

C. Incorrect. Attenuation processes are exponential, not logarithmic.

D. Incorrect. Logarithmic subtraction has no effect on contrast loss due to scatter. In fact, the presence of scattered radiation means that logarithmic subtraction cannot completely remove the effect of the mask image.

Question 11

WHat does the total patient dose depend on in DSA?

Answer 11

A higher II entrance dose results in a higher patient dose. The total patient dose depends upon the product of the patient dose per frame, the average number of frames per run and the average number of runs acquired during the examination. A complete DSA examination typically comprises many tens or hundreds of frames.

Question 12

What does the dynamic range depend on in DSA?

Answer 12

Dynamic range depends upon the section of the body imaged, the angiographic technique used and also the technical specification of the DSA image intensifier television (IITV) system.

The latter is strongly dependent upon the magnitude of the electronic noise introduced by the TV sensor. TV noise manifests itself most strongly in regions of the image corresponding to the sections of anatomy with particularly strong x-ray attenuation.

The dynamic range of a DSA system can be artificially enhanced by placing field equalisation or ‘bolus’ material over regions of particularly high x-ray intensity. An example of this is the placement of a long wedge-shaped aluminium filter between the legs in a DSA examination of the peripheral arteries.

Question 13

How is the nyquist frequency of a DSA system calculated?

Answer 13

Number of pixels/(2xinput field diameter)

Question 14

How can spatial resolution be improved in DSA?

Answer 14

Increasing the dimensions of the pixel array, say by upgrading the digital system specification from (1024 x 1024) to (2048 x 2048)
Selecting an II zoom field (with smaller II input field diameter)

Question 15

What other factors effect spatial resolution of DSA?

Answer 15

Factors affecting the resolution of a DSA imaging system:

The size of the x-ray focal spot and the degree of geometrical magnification
The resolution of the II tube
The resolution of the TV sensor and the bandwidth of the camera electronics
Possibly the resolution of the display device

Question 16

WHat is the typical spatial resolution in DSA?

Answer 16

The measured resolution of a large-field DSA system typically lies in the range 1.2 to 1.4 line pairs per mm. When an II zoom field is selected the resolution in DSA will improve by a factor of two to three.

Question 17

What are the 2 main sources of noise in DSA?

Answer 17

X-ray quantum noise manifests itself in regions of the image of high x-ray transmission.

TV camera electronic noise is added to the video signal and manifests itself in regions of low x-ray transmission.

X-ray quantum noise and TV camera noise have distinctive spatial imaging characteristics; the former is coarse grained while the latter is much finer grained.

The interplay of these two noise sources in differing regions of the image gives rise to the so-called ‘noise print effect’ which is familiar in DSA images.

Question 18

Why is the noise in the subtracted image bigger than either the mask or contast image in DSA?

Answer 18

i.e. the noise in the subtracted image is the square root of two (1.4) times greater than that in an individual image.

X-ray quantum noise and TV camera noise fluctuate randomly over space and time. Their contributions in acquired image frames are temporally uncorrelated. As a result the noise in the subtracted image σDSA is larger than that in either the mask image σM or the contrast medium enhanced (iodinated) image σI.

Question 19

WHat factors does the vessel SNR in DSA rely on?

Answer 19

Injected contrast concentration, diameter of the vessel, square root of dose contributing to the image

Question 20

What are the arguments for and against the use of CO2 as a contrast agent in DSA?

Answer 20

As a contrast agent, CO2 has certain advantages when imaging allergenic patients. However, as the subject x-ray contrast of CO2 is lower than for a full strength iodinated contrast agent, it may be necessary to use a higher radiation dose.

Question 21

What artefact does patient motion in DSA casue?

Answer 21

misregistration artefacts which can degrade the visibility of the iodinated vessels.

A movement of a structure or organ by less than a millimetre is sufficient to cause serious misregistration artefacts.

Clinical DSA systems include facilities to help to correct misregistration artefacts and salvage the procedure. Remedial techniques include pixel shifting and re-masking

Question 22

Regarding DSA imaging:
A. Quantum noise is halved by log subtraction
B. The typical operating dose per frame is approximately ten times that used in fluoroscopy
C. Voluntary or involuntary movement of the patient degrades image quality
D. Pixel shifting is used to correct for movement artefacts but at a cost of increased patient dose

Answer 22

A. Incorrect. Quantum noise is increased by a factor of √2 (about 1.4) by log subtraction.

B. Incorrect. The II entrance dose per frame in fluoroscopy typically falls in the range 10 to 40 nGy (0.01 to 0.04 μGy). The II entrance dose per frame in DSA is typically 1 to 30 μGy. Therefore the operating dose per frame in DSA is 25 to 3000 times that used in fluoroscopy.

C. Correct. Voluntary or involuntary movement of the patient produces misregistration artefacts which can degrade the visibility of iodinated vessels.

D. Incorrect. Pixel shifting does not increase patient dose.

Question 23

The SNR of a DSA image can be improved by:
A. Averaging the complete sequence of images following injection of contrast medium
B. Increasing the II entrance dose per frame
C. Diluting the contrast medium with saline
D. Adopting an intravenous approach

Answer 23

A. Incorrect. Many of the frames in the sequence will contain little or no contrast medium, so averaging the complete sequence will reduce the signal in the subtracted image. Frame averaging will reduce noise, but this is less effective in subtracted images because of the contribution of noise from the mask image. The reduction in noise will be outweighed by the reduction in signal, so the SNR will not improve. However, it may be possible to improve the SNR by averaging selected frames containing high concentrations of contrast.

B. Correct. The SNR is proportional to the square root of the II entrance dose per frame.

C. Incorrect. The SNR is proportional to the concentration of the contrast medium.

D. Incorrect. An IV approach will result in greater dilution of contrast medium and will therefore reduce the SNR.

Question 24

A realistic value for the spatial resolution of a DSA system with an II input diameter of 30 cm and a 1024x1024 pixel array is:

Answer 24

The Nyquist frequency for such a system is given by:

fN = 1024/(2 x 300)

fN = 1.7 lp/mm

In practice, the resolution will not reach this theoretical maximum value because it will be degraded by factors such as the size of the x-ray focal spot and the resolution of the II tube. Therefore a realistic value is about 1.4 lp/mm.

Question 25

WHat is bolus chasing?

Answer 25

The bolus chasing technique is a digital vascular imaging technique dedicated to the examination of the lower peripheral arteries.
A series of images of the peripheral vessels are acquired as the patient table moves continuously past the image receptor (or alternatively the imaging system moves past the patient). Real-time images are used to match the table translation speed with the progress of the contrast medium. This technique produces an overview of the complete lower peripheral vascular tree following injection of a single bolus of contrast medium

The bolus chasing technique is efficient in the use of:

Contrast medium
Radiation dose
Examination time

Question 26

WHat is rotational angiography?

Answer 26

Rotational angiography is a technique which is used to examine a general vascular site in greater detail.

While the C-arm is rapidly rotated around the site of interest, digital fluorography (DF) images are acquired.

The images are subsequently viewed as a movie loop, showing the vessels in rotation. They may be displayed as either basic greyscale or subtracted images

Question 27

What is 3D reconstruction angiography?

Answer 27

The set of DF images acquired during a rotational angiography sweep can often be transferred to a dedicated workstation to compute a 3D image, similar to volume CT imaging.

Surface-shaded display techniques are used to generate 3D representations of the vasculature that are anatomically more realistic than is available with rotational angiography.

The resulting 3D display can be viewed from many perspectives or can be rotated as it is being viewed. This can provide a clearer impression of the morphology and distribution of complex vascular segments.

Question 28

WHat is dual energy subtraction?

Answer 28

Dual energy subtraction uses the digital subtraction of images acquired at two different x-ray beam energies. This makes it possible to address two types of tissue, for example soft tissue and bone, and present them in separate images.

Question 29

How does DES angiography work?

Answer 29

Dual energy subtraction angiography algorithms exploit the discontinuous increase in x-ray absorption of iodine at energies straddling the K-absorption edge at 33 keV.

The discontinuity in the absorption coefficient means that an image using a photon energy above 33 keV is significantly more sensitive to the presence of the contrast medium than an image using photon energy below 33 keV

The DES angiography algorithm relies upon the acquisition of a series of high and low energy image pairs in rapid succession as the contrast medium flows through the vessels of interest.

Question 30

Which of the following propositions do you agree with:
A. Bolus chasing can be adapted for neurovascular imaging
B. Rotational angiography is the basis for 3D reconstruction angiography
C. Dual energy subtraction can discriminate between a calcified and a non-calcified nodule
D. Hybrid subtraction gets its name from the fusion of the DSA and temporal subtraction techniques

Answer 30

A. False. Bolus chasing is a means of imaging a large area using a single injection of contrast medium. It can only be applied when the bolus of contrast medium moves in a single, predictable direction. It is not applicable to neurovascular imaging.

B. True. The set of DF images acquired during a rotational angiography sweep may be transferred to a dedicated workstation to compute a 3D image.

C. True. In DES, a calcified lung nodule would appear in the bone image whereas a non-calcified nodule would appear in the soft tissue image.

D. False. Hybrid subtraction is a combination of dual energy and temporal subtraction techniques.