MI 237 - Unit 3

binary code, pixels, matrix, FOV

what are the building blocks of a digital image?

digital imaging

any process that produces an electronic image that can be viewed & manipulated on a computer

computer

electronics that can perform mathhematical computations

algorithms

preprogrammed set of instructions

in binary code

how does an algorithms operate?

inputs analog info. & converts it to binary code (discrete packages)

what is the main job of the computer?

digtial images made of numerical data

binary code

1

how many bytes are in 8 bits?

1,024

how many bytes are in 1 kilbytes?

pixel

smallest component of a matrix

picture element

what is another word for pixel?

a single brightness/gray shade level on a monitor

what is a pixel represented as?

by DEL hardware

in DR, how are the number of pixels determined?

by the laser diameter & scanning frequency

in CR, how are the number of pixels determined?

radiation intensity/remnant beam signal

every pixel is assigned a discrete numerical value in the binary bit system based on what?

in a martix

how are pixels arranged?

matrix

number of pixels

rows & columns (array) of pixels that make up the digital image

field of view (FOV)

overall dimension (size) of a matrix or portion of the IR that contains the relevant anatomic info.

size & meaurement

ex → 14 X 17

pixel size will decrease

spatial resolution will increase

if the FOV remains the same, but the matrix size increase, what else is affected?

inversely related

what is the relationship between the FOV to the matrix size?

directly proportional

what is the relationship of the pixel size to FOV?

FOV/matrix

how do u find the pixel size?

distance from the midpoint of one pixel/DEL to the midpoint of the adjacent pixel/DEL

pixel pitch

DEL pitch

what is another word for pixel pitch?

becomes smaller

if u decrease the size of the pixel, what happens to the pixel pitch?

bit depth

the number of available bits/gray shades a pixel can display

2^n

how to find the bit depth?

contrast resolution

if the bit depth increases, what also increases?

scanning, sampling, digitization

what is the correct order for digitizing an image?

scanning

analyzing the size of the latent image to format a matrix w/ a designated pixel size

occurs as the laser moves across the IP during readout

scans a designated number of lines based on cassette size

CR scanning

30%

how much should the CR cassette be exposed to complete the scanning process?

spatial resolution

in the CR scanning process, if u decrease the laser diameter, what increases on the image?

DR scanning

synonymous w/ collimation → (only the DEL reacting to radiation will be apart of image)

the incremental measurement of analog wave intensity

sampling

dictates the frequency of smapling must be greater than 2 times the frequency of the incoming signal

what is Nyquist Theorem?

if u increased smapling frequency = increased spatial resolution → better detail

what is the main point of Nyquist Theorem?

aliasing

insufficient sampling

morie effect → false lines across entire image

what does the aliasing cause?

low frequency

what causes aliasing/insufficient sampling?

quantization

assigns each sampled value a discrete numerical value

rounds pixel number for to nearest whole number/shade

on bit depth

what is quantization dependent on?

raw image which lacks contrast or has reduced contrast

before preprocessing

image that takes place in the computer

preprocessing

the image will be manipulated by techs & doctors after final image is displayed

post-preprocessing

field uniformity (flat fielding)

aims for consistent background shade/brightness

corrects inherent physical limitations

anode heel effect

response variations

what can cause a field unformity correction?

5

how many areas of the background does field unformity use to make the image uniform value?

one or more DELs stop working properly

results in area of abnormal signals

DEL drop-out (dead DEL)

small white square → no signal

what does a dead DEL look like on an image?

during preprocessing w/ kernal

when is a dead DEL corrected?

submatrix passed over an image multiple times & replaces pixel w/ average surrounding pixels

kernal

noise corrections

background noise inherent in all electronics

random distributions of photons

unwanted lower energies → under minimum field of signal

scatter reaching IR

what causes noise in image?

level of background noise identified & subtracted from all pixels values

noise subtraction

bar graph that counts all pixels sharing the same pixel value (brightness)

histogram analysis

number of pixels of intensity/energy

what is the vertical side of a histogram mean?

pixel values, amount of exposure signal, pixel shades, energy density

what does the horizontal side of a histogram mean?

the body part

what defines the pattern or shape of the histogram?

extreme ranges of strong intensities (dark/blacks)

the spike/tail of the histogram means what?

main lobe & direct exposure spike

most common

body parts that do NOT cover the IR

type 1 histogram

main lobe only

body parts that cover whole IR

type 2 histogram

type 1

what type of histogram would appear for an elbow?

type 2

what type of histogram would appear for a chest or abdomen?

spike light values, main lobe & direct exposure spike

barium or hardware

type 3 histogram

type 3

what type of histogram would appear for a knee replacement?

long gray scale → low contrast

a shallow slope would be consistent w/ what?

short gray scale → high contrast

a steep slope would be consistent w/ what?

histogram analysis

computer compares the acquired histogram to a stored histogram

exposure field of recognition

algorithms that identify the useful pixel values (VOI) w/in the histogram

subtraction

what is used to help identify the minimum & maximum data values for the exposure field recognition?

compares the count of one bin compared to the count of the next bin

subtraction algorithm

right to left

in what direction does the subtraction algorithm scan?

maximum & minimum

what intensities does the subtraction algoritm subtract?

VOI

what is remains after the subtraction of the minimum & maximum values?

scan for a pixel count exceeding a present threshold count

looks to predetermine landmarks

threshold algorithm

noise & background radiation → removal of lower energy pixels

what does the threshold algorithm help reduce?

exposure indicator

picks out median of histogram → only w/in the range of min & max

segementation/pattern recognition

exposure field recognition of CR

pattern recognition

identifies edge of exposure field on the cassette

rescaling

mainpulates the overall brightness & contrast level

compensates for most over &/or under exposures

acquired histogram shifted to more match ideal stored histogram

changes overall image brightness (intesnity of light)

increases or decreases ALL pixel values in histogram

moves histogram side to side

window leveling

window width

changes overall image contrast

accomplished w/ look up table

changes range (shape) of histogram → NOT height

look up table (LUT)

change pixel value in set increments starting w/ median pixel value

data/stored info. for every body part/projection (adjust contrast to preset desired appearance)

adjust new min & max luminance values to raw image rescaling the contrast

rounding

changing the severity of the rounding process during digitization affects the WIDTH of the histogram

either increases or decreases the number of gray scales

orginal pixel values

S-values

perfer pixel values used to entire constant image appearance

Q-values

histogram analysis errors

improper collimation

large tissue density (hardware)

improper location of anatomy (off-center, pt anatomy)

multiple exposures on CR plate (ghosting)

selecting wrong body part

w/ kernal

how do u correct a dead detector element (DEL)?

w/ noise subtraction

how do u correct random unwanted signal?

field unformity

how do u correct uneven background?

w/ rescaling

how do u correct incorrect brightness/contrast?

low attenuation, raw data, higher intensities

what does the movement of the histogram to the right indicate about the exposure index?

high attenuation, low intensities

what does the movement of the histogram to the left indicate about the exposure index?

brightness

window level = what?

contrast

window width & LUT = what?

continuous

what type of signal is an analog signal?

the computer is not in the binary code

why can a computer NOT store an analog signal exactly as it exists?

10

which set of sampling lines (5 or 10) more accurately represents the orginial signal?

more information

why is it that the more sampling lines are present, the closer to the original signal?

missed information

what can occur if the sampling frequency is too low?