Digital image processing-CT

Visual impact of irreversible compression on digital images

As compression ratios increase less storage space is required and transmission speeds increase but at the expense of image quality degradation ,Compression ratio of 8:1 or less visually acceptable

Types of image compression

lossless or reversible compression 2:1 or 3:1, no loss of information, doesn't involve the process of quantization but makes use of image transformation and encoding to provide a compressed image, Lossy or irreversible compression 10:1 or 50:1, some information lost, involves image transformation, quantization, and encoding, not used by radiologists in primary diagnosis due to legal concerns and incorrect diagnosis , At low compression ratios (8:1 or less) the image is still visually acceptable, compression tolerance, the maximum compression in which the decompressed image is acceptable for interpretation and aesthetics

Global operations

the entire image is used to compute the value of the pixel in the output image, a common global operation is Fourier domain processing(FT), uses filtering in the frequency domain rather than the space domain, can provide edge enhancement, image sharpening, and image restoration

Data acquisition

1st step- systematic method of collecting data from the patient, Attenuation data collected for imaging modality, Ct- x-Ray tube and digital image detectors do this, Output signal is an electrical signal(analog signal that varies continuously on time), Analog, signal must be converted into a digital signal for, computer processing using an ADC converter

Convolution(bone/lung window essentially)

A filter in the space domain in which the value of the output pixel is calculated by multiplying each input pixel by its corresponding weighting factor of the convolution kernel-basically a filter, This process is applied to all pixels in the input image, This is a post processing image reconstruction filter that can be applied to an image to view the scan in a different way to better see different anatomy

Matrix

A two dimensional array of numbers that make up a digital image, consists of columns(m) and rows(n) that define small square regions called picture elements(pixels), The matrix size is selected by the technologist(FOV), In CT we use 512x512

Voxels

A volume of tissue in the patient represented by pixels, Voxel information is converted into numerical values contained in the pixels, Numbers are assigned brightness levels, The higher numbers represent high signal intensity from the detectors and are shaded white(bright), The low numbers represent low signal intensity and are shaded dark(black).

Image processing hardware

ADC, Image storage, Image display, Image processor, Host computer, DAC

Spatial frequency filtering: high pass filtering

Also known as edge enhancement or sharpness, Intended to sharpen an input image in the spatial domain that appears blurred

Image synthesis overview

An advantage of digital image processing, Attenuation data is collected from the patient by the detectors, sent to the computer, and digitized by the ADC. Special reconstruction, algorithim(filtered back projection) is used to build up a digital CT image and us then converted to gray scale for viewing. Displayed images can be manipulated with digital imaging software to perform such processing as windowing, image reformatting, and 3D visualization

Local operations

An image processing operation in which the output image pixel value is determined from a small area of pixels around the corresponding input pixel, A group of pixels is used in the transformation calculation, *Spatial frequency filtering, if the brightness of an image changes rapidly with distance in one plane the image is said to have high spatial frequency, filtering can alter the image by sharpening,smoothing, blurring, noise reduction and feature extraction(edge enhancement and detection), two places to perform spatial frequency filtering, In the frequency domain(Fourier transform), In the spatial domain(pixel values/gray levels)

Image formations and representation analog images

Analog images are continuous images, Images formed when x-rays pass through patient and projected onto x-Ray film, Films processed in chemical solutions to render visible. These images formed by photochemical process, Images can be formed by photo electronic means, images may be represented as electrical(analog) signals that emerge from photo electrical device

Limitations of film based imaging

Been around since 1895 Used in all radiology modalities. Patient is exposed to predetermined amount of radiation that is needed to provide a diagnostic image. If image is under/over penetrated the patient would have to be exposed again, increasing dose. Not ideal way to perform radiation detection, image display and image archiving. Film-screen(x-Ray) can't show differences in tissue contrast less than 10%(contrast resolution). Optical range and contrast for film are fixed and limited(can't window/level them, have to retake). Stored in envelopes requires manual handling for archiving and retrieval.

Image digitization

Convert an analog image into numerical data for computer processing, Three steps, 1.scanning, results in a grid characterized by rows and columns, 2.sampling, measure the brightness of each pixel 3.quantitization, the brightness value of each sampled pixel is assigned a number, this results in a range of numbers(gray levels) with a precise location on the rectangular grid of pixels, results in a digital image, Gray scale=total number of gray levels, More gray levels means more shades of gray, An image can be composed of any number of gray levels, depending on bit depth

Analog-to-digital conversion

Converts analog to digital signal, Samples the analog signal at various times to measure its strength at different points, Sampling process then followed by quantization, Speed, the time taken to digitize the analog signal, Accuracy, the more samples taken the more accurate the representation of the digital image, aliasing-not enough samples are taken and the image isn't accurate after computer processing, Determines the number of levels or shades of gray represented in the image

Image processing hardware

Data acquisition, CT x-Ray tube and detectors. Digitizer, analog signal is converted into digital form. Image memory, digitized image is held for further processing. DAC, converts digital image to analog for viewing, Internal image processor, high speed processing of the input digital data. Host computer, primary component capable of performing several functions

Major components of generic digital imaging system

Data acquisition, Image processing, Display/storage/archivingImage communication

Image compression

Digital compression refers to using one or more of many, software and or hardware techniques to reduce information by removing unnecessary data, Remaining information then coded and either transmitted or stored in an archive or storage media Using decompression, the users equipment later decodes the information and fills in a representation of the data that was removed during compression

Image domains

Digital image processing can transform one image domain into another image domain, An image in the spatial location can be transformed into a spatial frequency domain image, The Fourier transform(FT) is used to perform this task, converts a function in the time domain to a function in the frequency domain, the inverse FT is used to transform an image in the frequency domain back to the spatial location domain for viewing by radiologist/technologist

Image processing: an essential tool for CT

Image post processing methods are intended to enhance the diagnostic interpretations skills of the radiologist, Examples of these include reformatting axial images to: coronal, sagittal or, oblique views, maximum or minimum intensity projections, curved reformatting, shaded surface display, virtual reality, image fusion or perfusion

Digital image processing

In image processing it's necessary to convert an input image into an output image, if both are analog this is referred to as analog processing, if both are discrete this is referred to as digital processing, When an analog image must be converted into digital data for computer input a digitizations system is required , CT is based on a reconstruction process whereby digital image, changed into a visible physical image, process-"a series of actions or operations leading to a desired result; thus, a series of actions or operations are performed upon an object to alter its form in a desired manner".

Geometric operations

Intended to modify the spatial position or orientation of the pixels in an image, Result in the scaling and sizing of images and image rotation and translation

Pixels(2D)

Make up the matrix, Generally square, Containsa number(characterized by the tissue imaged) that represents the brightness level. In CT, the numbers are related to the atomic number and mass density of the tissues, Pixel size can be calculated by: pixel size=FOV/matrix size The larger the matrix size the smaller the pixel size and the better the spatial resolution

Point operations

Most frequently used image processing technique, Most commonly used is gray-level mapping, uses a lookup table which plots the output and input gray levels against each other, changes the brightness of the images and results in the enhancement of the display image, results in a modification of the histogram of the pixel values

Bit depth

Number of bits per pixel, "K bits" implies that every pixel in the digital matrix MxN is represented by k binary digits, Because the binary number system uses the base of 2, k bits=2k, Each pixel will have 2k gray levels, In CT we use a bit depth of 12.

Image formation and representation digital images

Numerical representations or images of objects, Formation of any digital image requires digital computer, Any information that enters the computer must be converted into digital form(numbers). ADC converts continuous signals to discrete signals(digital data). Computer receives data and then does necessary processing. Result=digital and can be displayed as digital image, Output=analog image

Image display, storage, and communication

Output digital image must be converted into an analog signal before it can be displayed on the monitor using a DAC, Can be archived on magnetic data carriers and laser optical disks for retrospective viewing and manipulation, Can be communicated electronically through computer networks to remote sites

Image processing

Performed by digital computer , Takes input digital image, processes it to produce output digital image(using binary number system), ADC sends digital data for digital image processing by a digital computer, Accomplished by set of operations, techniques to the input image into an output image, these operations can reduce noise, enhance sharpness, or change the contrast

Spatial frequency filtering: low pass filtering

Smoothes the input image, The output image will appear blurred. Intended to reduce noise and the displayed brightness levels of the pixels,Image detail is compromised

Effects of digital image parameters on the appearance of digital image

The characteristics of a digital image can affect the appearance of the digital image(spatial resolution and density resolution), Matrix size, Bit depth

CT as a digital image processing system

The slice of the patient is divided into small regions called voxels because the dimension of depth(slice thickness) is added to the, pixel, The voxels are sampled when x-rays pass through them. The signal from the detector is in analog firm and must be converted into digital form before it can be sent to the computer for processing, The analog signal is quantized and changed into a digital array for input into the computer. The digital data is then sent through several processing algorithims so the output image can be displayed as a readable image

Characteristics of the digital image

The structure of a digital image can be described with respect to several characteristics or fundamental parameter. Matrix, Pixel, Voxel,Bit depth

Spatial frequency domains

Used in MRI, frequency refers to the number of cycles per unit length(the number of times the signal changes per unit length).

Spatial location domains=CT

Used in radiography and CT, The digital image is a numerical image arranged in such a manner thAt the location of each number in the image can be identified using the x-y coordinate system, x=columns(horizontal), y=rows(vertical), The first pixel in the upper left hand corner is always identified as 0,0

Spatial frequency processing: unsharp(blurred) masking

Uses the blurred image produced from the low pass filtering process and subtracts it from the original image- getting all blur out of the image