Computers and imaging :Computers and imaging Hardware: CPU central processing unit. Three primary functions: regulate system operation and perform computations, interact with memory to execute programs and store data, and coordinate the control of input and output devices.


Computers and imaging :Computers and imaging CPU acts as “brain” of the computer it retains information by storing it in memory Memory consists of two types: ROM: read only memory, RAM: random access memory. ROM is permanent memory, RAM can be changed and is lost when computer is turned off


Computers and imaging :Computers and imaging Input/Output devices compose the rest of the computer: display terminal, mouse, printers, modem, etc. Data storage comes in the form of magnetic disks: either floppy or hard disks. Storage capacities come in many ranges.


Computers and camera interface :Computers and camera interface Output from a scintillation camera detector provides three signals for gamma ray scintillation. X and Y signals represent the location of the gamma ray interaction in the detector Z signal measures the gamma ray energy.


Camera interface :Camera interface Three signals are analog or electronic pulses, that must be converted into numbers to be transferred into the computer. Analog to digital converters digitize X and Y signals.


Camera interface :Camera interface Imaging computers require certain environmental criteria’s, temperature, humidity, electromagnetic fields, power, and cleanliness. Changes in temperature, humidity, power fluctuations, and cleanliness can effect the final images.


Camera interface :Camera interface Imaging computers require some QC (quality control). Flood uniformity, and, resolution and linearity, are performed daily or weekly. Size, shape, and resolution of test images are evaluated for distortions and recorded to ensure proper image to storage capabilities.


Camera interface :Camera interface Software is a set of instructions or programs that provide control over the calculations and subsystems of the computer. Three principal types of software, systems software, programming software, and user software.


Image acquisitions. :Image acquisitions. Capturing images requires the conversion of analog x,y, and z to digital information. Each gamma event is recorded into a pixel and it is digitized. Scintillation event are stored in the proper pixel location until the acquisition time has elapsed, or until the selected amount of counts has been reached.


Image acquisition :Image acquisition Image density for each pixel is determined by number of gamma ray counts stored in each pixel. Image display is determined by the highest pixel count is assigned with the brightest intensity on the CRT display.(pg 110 fig:4-6, 4-7) Each pixel can store a limited number of counts, pixel overflow results when the number of counts are exceeded.


Image acquisition :Image acquisition Most computer systems allow for matrix selections of 64x64, 128x128, 256x256, 512x512. Static images contains enough counts to produce high spatial resolution images. Generally high matrix sizes are used. Dynamic images do not contain enough counts to warrant high resolution matrices, they are studies to record physiological information.


Image acquisitions and display :Image acquisitions and display Display screens usually allow images to be displayed as either black on white or white on black. Each pixel is assigned a gray scale value based on the number of counts. Images are displayed on the brightest pixel assigned to the maximum display intensity.


Image acquisitions and display :Image acquisitions and display Display intensities range (gray scale) from 0 to 255, maximum pixel count is assigned 255, all other pixels are assigned from 255-0 Relationship between the number of counts and display intensity is linear. Provides a uniform gray shading between all counts.


Image acquisitions and display :Image acquisitions and display Most computers allow for logarthimic and exponential relationship between pixel count and intensity Exponential suppress number of gray scales assigned to low count values while expanding the number of shades of gray assigned to higher count pixels, this reduces low count pixels thus removing background.


Image acquisitions and display :Image acquisitions and display Logarithmic assigns more gray scales or levels to low count pixels and compresses the number of shades of gray assigned to high count pixels, which enhances differences in low count densities. Color and background subtraction are other ways of enhancing images.


Image enhancement :Image enhancement Image smoothing performed to reduce noise from random effects of radionuclide counting. The simplest technique is to average counts given to pixels with that of surrounding pixels and replace center pixel with new value.


Image enhancement :Image enhancement Cinematic display is a dynamic study in which the images may be displayed in a continuous loop, similar to a movie. Dynamic studies, blood pool images, and SPECT images can be displayed in cinematic mode.


Image enhancement :Image enhancement Many nuclear images are acquired to derive quantitative information. Counts from a particular area can be extracted from the image by defining a region of interest (ROI) ROI allows to display or print the counts within that region, the number of pixels used, and the average count per pixel.


Image enhancement :Image enhancement ROI can be manipulated to either add or subtract regions.


Image enhancement :Image enhancement Quantitative information is derived by setting ROI’s and getting the counts derived from them. ROI counts in sequential images (dynamic studies) can be used to plot time activity curves. Curve displays may used in a wide variety of applications.


Image enhancement :Image enhancement Curves provide valuable information in evaluating the accumulation and washout of radiopharmaceuticals. Studies include: renal scans, gated heart studies, gastrointestinal studies, hepatobiliary ejection fraction studies,etc.


Image enhancement :Image enhancement Normalization is a concept in which a measurement has been brought to a standard. Ex: two images with different intensities (maximum counts) one low count and one high count are joined together, and the two are displayed with the same intensity. Most common application is to two ROI’S


Filtering :Filtering Image filtering to reduce noise depends on the information content of the image: counts, collimator, scatter, object distance, and background activity Frequency space filtering, there is always a trade-off between reducing noise and degrading resolution


Filtering :Filtering Images are converted to frequency space, objects or organs are represented as a group of low and middle frequencies. Noise is represented by high frequencies, which is separate from the frequencies that represent objects or organs of interest.


Filtering :Filtering In order to reduce noise in the frequency space, high frequency information must be reduced. Filter values for low, and middle frequencies retain the objects or organs of interest in the image. Filter values at higher frequencies drop to zero or small values, thus eliminating noise.


Filtering :Filtering Commonly used filters include: RAMP, Von Hann (hanning), Butterworth, Parzen, hamming, wiener, and Metz filters. All have same basic purpose to increase the amplitudes of objects frequencies and reduce the amplitudes of the high frequencies


Filtering :Filtering Ramp is a high pass filter: high resolution but also allows high frequency to pass thru, creating a noisy image. Von hann (hanning), hamming, butterworth, parzen are classified as low pass filters. Defined by variuos cutoff frequencies which allow low frequencies through.


Filtering :Filtering Low pass filters with a higher cutoff frequency will produce higher resolution but noisier image Filtering must balance high resolution against noise reduction. Butterworth filters is the best low pass filter. Contains both cutoff frequency parameter and order parameter. Which produces higher resolution images.


Filtering and SPECT reconstruction :Filtering and SPECT reconstruction Many filters used for frequency space filtering can also be used for filtering back-projection reconstruction Filtering for each type of patient to reduce noise and obtain high resolution is a matter of preference. Choices are between high resolution and a smooth image, yet increasing smoothing decreases resolution.


Filtering and SPECT reconstruction :Filtering and SPECT reconstruction Image filtering is done twice for SPECT imaging It is performed first as a pre-filtering function to remove noise from planar projections, then again during reconstruction of the image. The resultant image quality is influenced directly by image acquisition parameters: matrix size, number of stops(projections) counts per image, and filtering.


Filtering and SPECT reconstruction. :Filtering and SPECT reconstruction. Different exams use different filters with different cut-off frequencies Choosing the proper filter and cut-off frequencies will determine the resolution and smoothing of the final images. Single head vs Dual head cameras will produce less sensitivity and yield images with less counts and lower resolution.


Clinical applications :Clinical applications Most hospitals use the clinical applications programs supplied by the computer manufacturer. The manufacturer must provide evidence to the food and drug administration of clinical utility along with a description of the program. Each facility will have a slight variation in the use of the programs.


Workstations :Workstations Image viewing is generally handled on a multi-task workstation. Nuclear medicine networks can be set in many different configurations, depending on department needs.


WAKE UP!!!!!!!!!!!!!!!!! :WAKE UP!!!!!!!!!!!!!!!!! LIGHTS PLEASE………….