Digital Imaging Processing

1. Image aquisition 

2. Processing - Image enhancement; subject enhancement. - Image restoration; objective enhancement. Restoration's goal is to revert the image back to its original function prior to degredation. 

Image restoration can be mathematically quantified where as image enhancement cannot, or is more difficult to. 

3. Classification and diagnosis 

1. Spatial resolution 

2. Speed/frame rate

3. Signal to noise ratio

4. Contrast 

f(x, y)

f = pixel number indicating the grey level // (x, y) = pixel location 

a matrix of spatial coordinates with an assigned intensity level. each coordinate is indivdually assigned an intensity which is determined by the ADC bitdepth and the energy intensity level of the corrisponding sensor from the image detector. 

1. processes images to the optimal standard without the compromise of increasing dose 

2. images can be transfered between different doctors, hospitals and trsust with ease and can be transfered abroad 

3. allows for images to be processed, changed and stored 

spatial resolution is the smallest seperation between two objects that can be visible resolve as two seperate objects. 

• This determins the quality of the imaging system 
• Good spatial resolution: individual lines and spaces can be seen 
• units: line pairs per mm/dots per inch/ pixels per line 

this is the ratio between average signal and the stardard deviation of the background nose

•  this is a measure of sensitivity of a system 
• units: decibles 

• contrast is the seperation between the darkest and the brightest areas of an image 

• the more the contrast the wider the seperation between the two thus the more grey levels. 

• this is the time taken to aquire one image/single frame
• this is important in fluror - the quicker the speed the more the images 
• limtation: motion artefact 

• faster processing
• sensative sensors
• good spatail resolution
• less need to repeat 
• accomodating to high/low exposure 
• images can be easily transfered 
• less operator artefact 
• images can be enhanced or restored through post processing 
• electronic display and teleradiology 
• applicable to CT/MR/US 

dose creep 

• good spatial resolution 
• experienced radiologists/radiographers 

• chemical processing risk 
• film has a fixed non linear grey scale response 
• images need to be repeated if the exposure is not correct 
• digitisers required for teleradiography. 

1. illuminating source 

2. energy is either transmitted or relfected

3. the detctor records the intensity of the energy as a function of x, y; hence f(x, y)

- tranform incoming energy into a voltage (continous analogue signal) 

- the sensor material is tailored to do this with additonal electrical power 

- voltage converted to digital signal using ADC after 

• isloated 
• high spatial resolution 
• high quantum efficency 
• high signal to noise ratio 
• fast imaging acquision 

continous function (v) if a continous variable (t) 

= V(t) 

a discrete function Vk of a discrete variable tk with k (integer)

= V(tk)

high noise immunity 

ease of design 

less need to calibrate and maintain 

ease of diagnosis and repair 

fabrication

more reliable 

easily controlled by a computer 

easy to duplicate 

adjustable precision 

- low speed

- signal needs to be converted to communicate 

(A/D OR D/C) therefore more expensive and has less precision 

a rounding error between the analogue input voltage to the ADC digitised value 

Undersampling is when there is an insufficient sampling rate 

ALIASING 

this is when there is undersampling and there is not accurate representation of the analogue signal therefore results in distortion of the image. 

MOIRE PATTERNS 

opaqur lined patterns which are superimposed over similar patterns preventing distinguashable lines, thus causing artefact/distortion to an image. 

the sampling rate must be twice the maximum frequency measured in order for the analogue system to be completely represented 

sampling rate > 2 x fmax

line paris per mm = 1/width of 1 line pair 

this is the number of grey levels that are used to display the image 

this can be altered 

- ADC with higher bitdepth 

- Higher sampling frequency/pixel density 

the spaital resolution also decreases

this becomes evident when images are magnified - not always heavily effected on diagnostic quality 

focusing optics can be a limitation as the mechanical motions in the x and y plane are very precise 

the time taken to scan is long 

the number of elements/sensors in a *****

the density of elements/sensors in the array

high sampling rate and quantisation 

N x M x K 

these values are arranged into colomns (N) and rows (M) and this is done by a header file.

Bit depth determins the discrete grey level (k) for each pixel 

the number of samples in the vertical or horizontal direction

the bitdepth 

required for a good approximation?

A HIGH AMOUNT

A HIGH AMOUNT

A HIGH AMOUNT

the greater these parameters are the better the image as image quality is dependent on the number of pixels and number of grey levels

This is thhe smallest discerinble change in grey level - l 

The grey level is dertermined by the ADC

the spatial resolution 

files which receive the binary information from the ADC and organise them into specific locations; M and N 

• numer of pixels in the array
• dimensions of the image
• bitdepth
• line of binary defining the grey level of each pixel 

• support the distribution of medical images 
• allows for transmission of data between DICOM devices 
• allows for intergation of scnaners, serves worksstations and network hardware from different manufacturers into archiving systems 
• can be compatible with different diviced which have a DICOM comformage statement 

Digital Imaging and Communications in Medicine made by NEMA

allow for the distribution of medical images and is the stanard 

files that contain; patient name, machine ID, date 

only contain one attribute of pixel containing data (single images or miulti-frames)

Picture Archiving + Communication Systems 

1. Hard copy replacement

2. Remote access

3. Electronic integration 

4. Radiology workflow management 

the control of the amount of light enetring the eye by changing the size ofthe pupil. brightness adaptation can't opperate over different light ranges simultaneoulty.

small ranges of brightnesss can be discriminitaed simultaneously compared to total adaptation range. 

aka

brightness adaptation cant operate over a large change of light ranges simultaneously, however small ranges of brightness levels can be discriminated simultaneously. 

the ciliary body muscle relaxes -> lens flattens -> thinner lens -> longer focal length (17mm) -> lower referaction 

ciliary body muscle contracts -> thicker lens -> shorter focal length -> high refraction 

refracted power = 1/focal length 

• preceived intensity 
• logarithmic functions of light 
• lower light intensity 
• higher light intensity 

Webber found that the human eye can detect 12-24 intensities 

however this experiment had only one target source, a dot. An x-ray iage is a whole image where all of it needs to be assesed. As an x-ray has more than 24 shades the eye is still ablew to adapt to this through continous movement of the pupil. The eye therefore contously adjusts the brightness adaptation and is therefore able to see more shades of grey. 

it is important to ensure that the background lighting of the viewing environemtn is constant as it can cause changes to brightness adaptation. Optimal viewing would be dim lighting to ensure that there is a suitable light level. If there are changes to the background lighting it can cause the light level to be at a position where pathology can be missed if it is above the light level, causing it to appear white or below the light level, causing it to appear black. 

the ability of the eye to discriminate different intensity levels at given adaptation levels. In darker or brighter rooms some intensity levels may or may not be discriminated.