image data processing / reconstruction

Question 1

Define Interpolation? 2

Answer 1

1. Mathematical technique to estimate the value of a function from known values on either side of the function
2. Used in the reconstruction process of helical data

Question 2

Explain how filtered back projection can be used to render images? 5

Answer 2

1. Simple back projection produces blurred transaxial images.
2. Projections need to be modified before being back projected by using filters also known as kernals. i.e. smoother for SOFT tissue and sharper for high resolution imaging such as bone
3. Fourier transforms converts spatial image into the frequency domain, which is then applied to a high pass filter to supress the low frequencies. Hence, accentuating the high frequencies (detail and sharp edges) and reducing the blurring rendered by simple back projection.
4. The spectrum is inverse fourier transformed then back projected to render a shaper image.

Question 3

fill in the blanks 11

Answer 3

Question 4

FILL IN THE GAPS

Answer 4

Question 5

Describe how iterative reconstruction works highlighting the pros and cons 12

Answer 5

Advantages

1. able to handle the physics of projection/ back projection process
2. better handling of beam hardening and proton starvation
3. better handling of noise
4. can be used to lower the patient dose as a result of less noise.

Disadvantages

1. requires high computational power
2. images can appear waxy/ cartoon like hence reporting staff hesitent to change practice considering influence on diagnostic accuracy.
3. Slow reconstruction times but changing
4. Too many iterations can amplify noise

how it works

1. start with an assumed imaged and compute projections. Done by modelling the image system, true object and noise).
2. Sometimes the iterative process starts off the FBP data
3. This is compared with measured projection.
4. Review the assumption and repeat as part of a cycle until assumed and measured values are within acceptable limits.