CT data acquisition/ image reconstruction

data acquisition

method by which a patient is scanned to obtain enough data for image reconstruction

basic elements for data acquisition

beam geometry

components (physical devices)

beam geometry

size, shape and motion of the beam and its path

components

physical devices

pre patient collimator

shape the beam

detectors

measure the beam transmission through patient

ADC

convert info into digital data for input in computer

steps of CT data acquisition

1. The tube and detector are always in alignment​

2. The tube and detector scan the patient to collect a large # of transmission measurements ​

3. The beam is shaped by a special filter (bowtie filter)​

4. The beam is collimated to pass through only the slice of interest (pre- patient collimator)​

5. The beam is attenuated by the patient, passes through the post patient collimator, and the detector measures the transmitted photons​

6. The detector converts the photons into an electrical signal​

7. The signals are converted by the ADC​

8. The digital data is sent to the computer for image reconstruction where mathematical equations occur (algorithms)

methods of data acquisition

localize

conventional/ serial CT scan

helical, spiral, or volumetric CT

localizer scan (scout)

stationary tube/ patient table motion

provides an image of superimposed tissue

large field of view

allows alignment of cross-sectional slices with specific structures

conventional/ serial CT scan

tube rotates around the patient

table remains stationary

raw data

data measured in each projection

helical/ spiral/ volumetric CT scan

continuous data collection through multiple projections during continuous patient translation through the gantry

*needed for 3D image reconstruction

requirement for helical scan

scanner must be of continuous rotation

contain high heat capacity tube

rapid colling capacity

helical scan raw data

info must be divided into individual cross-sectional slices

mathematical interpolation must be run to divide the different samples form the different planes and raw data from different slices

advantages of helical CT scan

Complete organs may be scanned in one breath hold​

Many slices acquired at a time​

Less chance of mis-registration with inconsistent breath holding​

Smaller amount of contrast needed​

3-D reconstruction

multislice detector array

requires several parallel detector arrays that contain thousands of individual detectors

requires a fast large capacity computer

algorithm

set of rules or directions for getting specific output from specific input

fourier transformation

primary mathematical method used in CT for image reconstruction

takes complex data and arranges it into simpler and more useful forms

interpolation

the projecting of raw data between two known values

used to create wide variety of sections that are reconstructions of data

extrapolation

the projecting of raw data beyond the range of known values

retrospective reconstruction

saving the raw data from a scan

allows for post processing manipulation

filter back projection

filter refers to mathematical function

images form all projections placed together to form one image

image produced is not very sharp

convolution

process of applying filtration to the data

convolution

Process of modifying pixel values by a mathematical formula through a filter function​

mask

overlaps acquired data to reconstruct the image

removes star-like blurs from black projection

deconvolution

process of returning the pixel values to their original level by the reverse process

multiplanar reformation (MPR)

Post processing technique performed on image data which produces new slice from a set of CT scans​

Utilized for images in planes that ​would otherwise be difficult or ​impossible to acquire​

multiplanar reformation

software that allows CT to show an entire volume in one image 3D

**time consuming

maximum intensity projection (MIP)

Simplest form of three dimensional imaging​

Reconstruction can be done quickly​

Good for showing vasculature from surrounding tissue

Only uses 10% of the data points to create a three dimensional image​

shade surface display (SSD)

Does not generate images of cross-sections of anatomy but images the surface of the anatomical structure​

Stack transverse slices to form volume of anatomical data (contiguous slices necessary as to not miss any data)​

Reconstruction process dependent on separating different tissue types in the scanned images​

applications for SSD

Vessel display, soft tissue, other viscera​

Relationships between vasculature and viscera (thrombus or calcifications)​

Surface and internal detail of anatomy​​

Benefits for orthopedic and craniofacial surgery, neurosurgery, and radiation therapy​

shaded volume display (SVD)

utilizes 3D semitransparent representation

process utilizes all voxels that contribute to the image

shows multiple tissues and their relationship to one another