Image Processing

Spatial resolution

Is related to the sharpness of the image

Voxel

3D

Pixel

2D

Higher Pixel, Higher Spatial Resolution,

Increase Sharpness

Contrast resolution

Is a measure of the ability to distinguish small differences in intensity

Grayscale: 1 pixel is equals to

8 bits

Colored: 1 pixel is equals to

24 bits

Temporal resolution

Is a measure of the time needed to create an image

Image Processing

Aka digital image manipulation

Digital image processing

Use of a digital computer to process digital images through an algorithm

Image Preprocessing

Modifies and prepares pixel values to produce a form suitable for further processing

Image Preprocessing

Depends on modality and corrects system irregularities (light detection efficiency, dead pixels, dark noise)

Image Preprocessing

Preparation for succeeding processing

Image Preprocessing

Original Data/Raw Data/Raw Image

Branches of Preprocessing

Image enhancement and image restoration

Image Enhancement

Suppresses distortions

Image Restoration

Enhances important image features

1st Generation

Skin Film Radiography (SFR) and basic computer techniques (histograms, window width/level)

1st Generation

Higher frequencies adjust small structure visibility and noise

Edge Enhancement

Image filter enhancing edge contrast for sharpness (acutance)

2nd Generation

More complex algorithms, includes multiscale processing

2nd Generation

Enhancing only the right structures for optimal diagnostic view

2nd Generation

Capable of removing unwanted structures

3rd Generation

Goal is to eliminate need for user (RT) input

3rd Generation

Automatic body-part recognition, collimator recognition, application-specific processing

Image Filtering

Changing image appearance by altering pixel colors

Image Filtering

Increases contrast and adds special effects

Image Segmentation

Divides image into regions of similar properties (gray level, color, texture, brightness, contrast)

Image Segmentation

Assigns labels to pixels (2D) or voxels (3D) based on similar characteristics

Medical Imaging Use of Segmentation

Essential for quantification and 3D visualization of relevant structures

Feature Extraction

Transforms raw data into numerical features while preserving original information

Feature Extraction

Yields better results than applying machine learning directly to raw data

Feature Extraction

Removes part of data to evaluate further