TRIGGER 5 INFORMATICS

Preprocessing in radiography

involves a series of techniques used to improve the quality and usability of radiographic images for analysis and diagnosis.

Preprocessing in radiography

These techniques aim to reduce noise, enhance contrast, and remove artifacts that may hinder accurate interpretation.

denoising, contrast enhancement, background removal, and sometimes, image registration.

Common preprocessing steps:

Improve image quality

Enhance interpretability

Facilitate analysis

Preprocessing in radiography helps to:

Improve image quality

By reducing noise and artifacts, preprocessing makes the images clearer and easier to interpret.

Enhance interpretability

By improving contrast and removing irrelevant background elements, preprocessing highlights the relevant features for diagnosis.

Facilitate analysis

Preprocessing ensures that the images are in a suitable format for further analysis, whether by radiologists or by automated image analysis tools.

Image reconstruction

refers to the process of converting raw data collected by an imaging system into a visual image that can be interpreted by a radiologist. This is especially important in advanced imaging techniques like computed tomography (CT), but it also applies in simpler forms to digital radiography.

Digital radiography (DR) or computed radiography (CR)

the system captures raw X-ray data as electrical signals.

Image reconstruction

processes these signals to create a 2D image of the internal structures.

CT Scan

In _________, image reconstruction involves converting multiple X-ray measurements taken from different angles into cross-sectional images (slices) using complex mathematical algorithms.

Image reconstruction

In CT Scan, ____________ involves converting multiple X-ray measurements taken from different angles into cross-sectional images (slices) using complex mathematical algorithms.

1. Analog to Digital Conversion (DR/CR)

2. Filtered Back Projection (FBP)

3. Iterative Reconstruction (IR)

4. Fourier Transform Methods

Types of Image Reconstruction:

Analog to Digital Conversion (DR/CR)

Converts raw electrical signals from detectors into a grayscale digital image. Includes preprocessing and basic reconstruction of intensity patterns.

Filtered Back Projection (FBP)

A fast and commonly used algorithm that uses mathematical filters to reconstruct images from projections.

Iterative Reconstruction (IR)

A more advanced method that improves image quality and reduces noise or radiation dose.

Fourier Transform Methods

Uses frequency domain data to reconstruct images, sometimes used in MRI and CT.

Filtered Back Projection (FBP)

is a CT image reconstruction algorithm that combines filtering with back projection to create images from X-ray projections.

2 Main Steps in Filtered Back Projection (FBP):

1. Filtering the projection data to remove blurring.

2. Back-projecting the filtered data to reconstruct the image.

FBP

is computationally efficient but can result in noisy images.

Iterative reconstruction methods (IR)

offer potentially reduced noise and lower radiation dose.

Background Removal

is the process of eliminating unwanted or non-diagnostic background signals from an X-ray image to improve clarity and focus on the relevant anatomical structures.

Example of Background Removal:

-Unexposed or minimally exposed areas (like corners or borders)

-Artifacts from the detector (such as electronic noise or grid patterns)

-Structures outside the region of interest (e.g., bed frames, patient table edges)

-Light shading or ghosting effects

These background elements can reduce image contrast, obscure important details, or

distract radiologists from interpreting the image accurately.

Digital Subtraction Angiography (DSA)

Digital subtraction, in the context of medical imaging, specifically refers to ____________.

Digital subtraction

in the context of medical imaging, specifically refers to Digital Subtraction Angiography (DSA).

Histogram Equalization or Thresholding

are image processing techniques used for enhancing contrast and simplifying images.

Histogram equalization

redistributes pixel values to spread them across the entire intensity range, while thresholding converts grayscale images into binary images by setting a specific intensity value as a boundary.

Edge Detection and Cropping

helps identify boundaries of objects in medical images, while cropping removes unnecessary portions of the image.

Canny edge detection

Edge detection algorithms like _________ can pinpoint irregularities, aiding in tumor detection and

treatment planning.

Edge detection algorithms

like Canny edge detection can pinpoint irregularities, aiding in tumor detection and treatment planning.

Cropping

focuses on specific regions of interest, enhancing the clarity and analysis of certain anatomical structures, such as lungs or the brain.

Flat-Field Correction

a technique used to reduce image artifacts caused by non-uniformities in the X-ray beam or the detector itself.

Flat-Field Correction

These artifacts can manifest as shading, inconsistencies in pixel response, or variations in intensity across the image.

Flat-Field Correction

aims to normalize the image by removing these systematic defects, ensuring a more consistent and accurate representation of the object being imaged.

Noise removal algorithm

the process of removing or reducing the noise from the image.

Noise removal algorithm

reduce or remove the visibility of noise by smoothing the entire image leaving areas near contrast boundaries. But these methods can obscure fine, low contrast details.

Noise reduction

very important task in image processing because of the need for image analysis. It is important to preserve the details of the image when removing the noise so that the edges and edges of the objects remain clear.

low- pass (smoothing)

To remove noise, ____________ filters are applied to the image.

remove noise

To ___________, low- pass (smoothing) filters are applied to the image.

Noise pixels

are like edges and lines in that they stand out from their neighbors, and because of this similarity, removing noise in an image also results in removing, or at least blurring, edges and lines.

Noise in radiography/ Noise

refers to unwanted variations in an image that don't represent the subject's anatomy. It can be caused by

various factors and can reduce image quality and make it harder to interpret.

1. Quantum Mottle

2. Electronic Noise

3. Structured Noise

What causes Noise?

quantum mottle

The most common type of noise in radiography is

____________ , which is a result of the statistical fluctuations in the number of detected X-ray photons.

Quantum mottle

which is a result of the statistical fluctuations in the number of detected X-ray photons.

Quantum mottle

This is the most prevalent type of noise in X-ray imaging, including plain film radiography, mammography, and CT. It arises from the random nature of X-ray photon interactions with the detector.

more X-rays

The ___________ used to create the image, the less quantum mottle there will be.

less quantum mottle

The more X-rays used to create the image, the _____________ there will be.

Electronic Noise

In digital radiography, electronic noise can arise from various sources, including the detector itself and the analog-to-digital converter.

Structured Noise

unwanted, repeating patterns or artifacts in an X-ray image that resemble anatomical structures or consistent patterns, making it harder to distinguish real anatomy from false features.

Structured Noise

is not random—it has a pattern, which can be especially misleading in diagnostic.

Salt-and-pepper noise

also known as impulse noise, is a type of image degradation characterized by sparsely occurring black and white pixels, appearing as scattered dots on an image.

Impulse noise

Salt-and-pepper noise also known as________, is a type of image degradation characterized by sparsely occurring black and white pixels, appearing as scattered dots on an image.

Salt-and-pepper noise

It's a common problem in radiographic images, potentially impacting diagnostic accuracy.

Speckle noise

often associated with ultrasound, is a granular, textured appearance in the image due to the interference of echoes from multiple scattering points.

Speckle noise

This interference creates a pattern that can obscure fine details and reduce the contrast between structures.

Poisson noise

a type of random noise in digital radiography, arises from the statistical nature of X-ray photon detection.

Poisson noise

It's particularly prevalent in low-dose imaging, where the number of photons incident on the detector is limited.

Gaussian noise

a type of random noise following a normal distribution, can impact the quality of radiographic images, especially in digital radiography and other imaging modalities like CT and MRI.

Image compression in radiography

is the process of reducing the size of digital radiographic image files while maintaining as much diagnostic image quality as possible. This is done to save storage space, improve image transmission speed, and support efficient archiving and sharing—especially over PACS (Picture Archiving and Communication Systems).

1. Lossless compression

2. Lossy compression

2 types of image compression:

Lossless compression

-Preserves all original data

-Every bit of information in the original image is retained after compression and decompression.

-No quality degradation

-The compressed image is identical to the original, ensuring a perfect copy.

Lossy compression

-Reduces file size by removing some data

-Less critical or redundant information is discarded to achieve smaller file sizes.

-Potential for quality loss

-The compressed image may appear slightly different from the original, especially at high compression levels.