A Window into the Human Body
Image Quality and Visibility
The Five Image Quality Characteristics
Characteristics and Limitations of Medical Imaging Technology
Effect of Equipment Operation on Image Quality
An Extension of Human Vision
The First Test of Our Vision
Contrast Sensitivity
Contrast Transfer in Medical Imaging
Contrast Sensitivity and Visibility
Factors Affecting Contrast Sensitivity
Contrast Sensitivity Illustrated
Image Contrast
Measuring Object Contrast
Physical Contrast in X-ray Imaging
Matching Contrast Sensitivity to Physical Contrast
Physical Contrast in Ultrasound
Contrast Sensitivity in Radionuclide Imaging
Physical Contrast in Magnetic Resonance Imaging
Blurring and Visibility of Detail

The medical image is a window into the human body.

It is formed by the imaging modalities that use various forms of radiation and energy to open the body to visualization.

Image quality is the overall characteristic of a medical image that determines which objects and structures within a body are visible.
The general objective of most imaging procedures is to visualize the various anatomical structures and any signs of pathology if they are present.
 

Image quality, and visibility of the internal body, is determined by a combination of five more specific image characteristics.  They are:

• Contrast Sensitivity

• Blurring

• Visual Noise

• Artifacts

• Spatial (Geometric) Characteristics

Generally, the factors that affect image quality for any specific imaging method have a direct effect on one or more of these specific characteristics.  A knowledge of these relationships is necessary for:

• Understanding the limitations to visibility

• Evaluating image quality

• Setting up imaging protocols for specific clinical procedures and objectives.

The first thing that determines the characteristics and quality of a medical image is the imaging method or technology used to produce the image.

For medical imaging we have a choice of:

• Radiography (Including Mammography)

• Fluoroscopy

• Computed Tomography

• Ultrasound Imaging

• Magnetic Resonance Imaging

• Radionuclide Imaging (Planar scan, SPECT, PET)

The general principles and the technology of each of these methods determine the image characteristics.
That is why a specific method is selected to visualize the different clinical conditions.

For each imaging method (CT is illustrated here) the specific image quality characteristics are determined by how the equipment is operated.

For each imaging method there are certain technique or protocol factors that must be adjusted by the operator before performing the imaging procedure.

It is the selected values of these factors that determine the specific image characteristics.  For most imaging methods there is a range of image quality within which the equipment can be operated.

The actual quality within this range is determined by how the equipment is operated.

That raises an interesting question.... If the image quality can be adjusted, why not always adjust for maximum quality and visibility?

Answer.... In medical imaging there are many trade-offs and compromises that must be considered.

• Many times when a protocol factor is changed to improve on image characteristic, one of the other characteristics will be reduced in quality.

• Also, in x-ray imaging there is a strong relationship between image quality and radiation dose to the patient.

That is why the technique factors and imaging protocol for each image acquisition should be selected to produce an optimized procedure, based on the specific clinical requirements for visibility.

An optimized protocol produces a balance among the image characteristics (for example, blurring and noise) and uses the radiation dose that is necessary to produce the required image quality.
 

Medical imaging technology is actually an extension of our human vision.

Just as we use the microscope to view small things, and the telescope and television to view things at a distance, medical imaging equipment extends our vision into the normally invisible regions of the human body.

This is a useful concept because we can use some of the same characteristics used to evaluate human vision to evaluate a medical imaging procedure.

Now, Let's test our vision....

The test is to look at the chart and read as far up the alphabet that you can.

Can you see the G, the H, the I? Make a note of your test results.

Before we move on, let's notice the characteristic that is changing as we go up the alphabet.

It is the CONTRAST of the letters.  The letter A has a very high contrast in relationship to its background, and is very visible and easy to see.

As we read up the alphabet, it is the contrast of the letters that is decreasing, and reducing their visibility.

What characteristic of our vision have we just tested?

We have just tested the CONTRAST SENSITIVITY of our visual system.

Contrast sensitivity is the characteristic of an imaging system that determines the lowest contrast object that can be visualized.

It is one of the most significant characteristics of all medical imaging methods.

In medical imaging it is often necessary to visualize objects and structures in the body that themselves have very low physical contrast.  Soft tissue masses surrounded by normal soft tissue are good examples.

The function of all medical imaging procedures is to:

• Convert physical contrast into visual contrast.

• Transfer the contrast from the body to the image.

The relationship of the visual contrast to the physical contrast is determined by the contrast sensitivity of the process.

Contrast sensitivity is generally expressed as the lowest physical contrast that produces the necessary image contrast for visibility.

Contrast resolution is an alternate term that is sometimes used for this general characteristic  of a medical imaging process.

Visibility of anatomy and signs of pathology depend on the contrast that is produced and displayed in the image.

When the contrast sensitivity is low the result will be low image contrast and low visibility.

We cannot always detect low contrast sensitivity by just looking at an image.  We might have an image that appears to have high visual contrast.  Such an image might be showing some of the high contrast objects like, like bones and bullets, but not showing low contrast objects like soft tissue masses.

Contrast sensitivity, that is the ability to convert and transfer contrast, is determined by two factors:

• Inherent characteristics of the specific imaging method.

• Selection of imaging technique and protocol factors.

Each imaging modality has contrast sensitivity characteristics that relate to the type of physical contrast it can image and the lowest physical contrast that can be visualized.

The contrast sensitivity that is actually achieved then depends on the operating factors used for a specific procedure.

Here we can compare two imaging procedures, one with a low and the other with a higher contrast sensitivity.

As shown on the left, objects in the body have a wide range of physical contrast.

Bones, barium, and bullets (the three "B"s) have very high physical contrast compared to surrounding soft tissues.

Most soft tissue objects, such as cancers, usually have a very low physical difference (contrast) relative to the surrounding tissue.

When imaged with a low contrast sensitivity, only the high contrast objects are visible.

High contrast sensitivity is required to visualize the many low contrast structures, objects, and signs of pathology in the body.

The objective is not to always have the highest possible contrast in an image.  If there is high contrast between areas in an image (area contrast) visibility might be reduced in the very light and dark areas. This is especially true for images recorded on film because the contrast sensitivity of film is reduced in the light (low exposure) and dark (high exposure) areas.

Very low image contrast, as shown on the left, usually results in low contrast sensitivity and low visibility.

In most cases, there is usually an optimum image contrast that gives the best visibility of objects throughout the image.

Optimum contrast is obtained by the right combination of film type, film processing, and level of exposure.

In digital images, optimum contrast is obtained with the appropriate digital processing and adjustment of display factors, especially the windowing.

The contrast of individual objects within an image can be measured and given a numerical value.

The contrast of an object is expressed relative to its surrounding background.  That is what determines its visibility.

For an image recorded on film, the contrast of a specific object is the difference in the film density (darkness) values.  These can be measured with a densitometer.

Note: Measuring object contrast in a test film with a densitometer is a routine function in mammography quality assurance procedures.

For displayed images the contrast of a specific object is the ratio of the brightness of the object to the brightness of the background.

Objects and structures in the body must have some form of physical contrast in order to produce visible contrast in an image.

Physical contrast is the difference in some physical characteristic of the tissues or other materials in the body.  The physical characteristics that produce visible object contrast is different for the different imaging modalities.

For x-ray imaging, including CT, the two sources of physical contrast are:

• Differences in density

• Differences in atomic number (Z)

 Radiographs and CT images are generally images showing differences in density within the body.
Because of the air in the lungs there is a large variation in density in the chest which makes it a natural for x-ray imaging.

CT images also display differences in density.  However, CT has a very high contrast sensitivity compared to radiography and can display relatively small differences in tissue density, as in the head.

When there is not a sufficient difference in density within the body to produce visible contrast (like between fluids and soft tissue) compounds  of either iodine or barium can be administered to enhance contrast.

Iodine and barium are good x-ray absorbers, not so much because of their density, but because of their atomic numbers (atomic size). As described in later modules the arrangement of the electrons (their binding energies) in these two elements makes them especially good absorbers and sources of contrast. 

The contrast sensitivity of an imaging procedure must be adjusted according to the the amount of physical contrast in a specific anatomical region.

A radiographic technique that has low contrast sensitivity is used for chest imaging because of the high level of physical contrast (variation in density) that is present.

Mammography requires a technique with high contrast sensitivity because of the very low physical contrast in the breast.

There are several factors that determine the contrast sensitivity of a radiographic/mammographic procedure.  The major factor is the selected KV value as described in later modules.

In ultrasound imaging there are two principle source of physical contrast that can be imaged:

• Echo sites produced by differences in acoustic impedance of the tissues.

• Motion, usually of red blood cells

Different modes are used to image the two different types of physical contrast.

The B (Brightness) mode displays the intensity of reflections or echoes as image brightness.  A bright spot indicates a structure surface or boundary  that is producing strong echoes (because of a large difference in acoustic impedance).

The Doppler mode displays flowing blood.

Physical contrast in all types of radionuclide imaging (including SPECT and PET) is the variation of radioactivity throughout a body section.

Physical contrast depends on the differential uptake and distributions associated with the various physiological functions and pathological conditions.

The imaging equipment must be adjusted to maximize the contrast sensitivity to the specific radionuclide that is being used.

The critical adjustment is that of the energy window in the Pulse Height Analyzer (PHA) of the gamma camera.
It selects the specific radiation and radionuclide that will be visible.

MRI can produce visible images of several different physical characteristics as illustrated here.

Most imaging is based on the three magnetic (physical) characteristics of tissue:

• Proton Density

• T1

• T2

The contrast sensitivity to a specific characteristic is set by the selected protocol factor values.

Three forms of fluid movement are other physical characteristics that can be sources of contrast in MRI.  The visualization of each requires special imaging techniques.

Compare the two images shown here.  What difference do you see?

Would you agree that the image on the left is more blurred?

Now let's consider the effect of this blurring on visibility.

In the image on the right, with much less blurring, we have much better visibility of the anatomical details.

It is the blurring in medical images that limits visibility of detail.

There is some amount of blurring in all medical images.  A knowledge of this blurring effect and how to control it is necessary for optimizing imaging procedures to insure appropriate visibility.

This test chart can be used to evaluate visibility of detail.

The letters and numbers become smaller as we move down the chart, this represents increasing detail.

Most of us with normal or corrected vision can read all rows.

What would we see (or not see) if we had very blurred vision?

It is not your vision that is the problem!

This image has been blurred to demonstrate a very important concept.

Blurring reduces the visibility of small objects and detail.

This applies to all forms of medical imaging.

Objects in the body cover a wide range of sizes.

 The smaller objects represent increased anatomical detail.

Here we have arranged a series of objects by size or detail.

We will now use this to observe the effects of blurring.

In this image with relatively little blur, we can see all of the objects.

Now, let's blur the image.

These images show the effect of blurring.

Notice that even in the blurred images, the large objects are visible.

Blurring places a limit on the smallest objects that can be seen.

Blurring limits visibility of detail.

There are three specific effects of blurring.

• As we have observed, blurring reduces visibility of anatomical detail.  That is what is significant in clinical imaging.

• Blurring reduces the perceived "sharpness" of structures and objects in an image.  The term unsharpness is sometimes used as alternate name for blur.

• Blurring reduces a characteristic that is known as spatial resolution.  Resolution is an effect of blur that can be measured relatively easily and is used to evaluate and specify the blurring characteristics of imaging systems and imaging components.
  This process is covered in a later module.
  Resolution is expressed in terms of the number of line-pairs in a space of one mm that are visible. Increasing LP/mm generally relates to increasing detail.
  Therefore, a high spatial resolution indicates high (good) visibility of anatomical detail.

The clinical significance of the blurring that occurs in all medical imaging procedures is that it places a limit on the smallest object (anatomical detail, sign, etc) that will be visible.

The amount of blurring (and visibility of detail)  in a specific imaging procedure is determined by a combination of two factors:

• The blurring characteristics of each specific imaging modality.

• The selection of technique and protocol factors for each imaging procedure.

There is a range of blur values for each imaging modality as illustrated here.  Within that range, the blurring in a specific imaging procedure is determined by the selection of technique and protocol factors.

The difference in the amount of blurring from one modality to another is because of the different principles and methods of image formation and the design characteristics of the equipment.

The scale shows the size of the blur and the associated maximum spatial resolution which relates to visibility of detail.

Notice that of all of the modalities, radiography produces images with the least blurring and therefore the best visibility of detail.

Mammography is the special radiographic procedure that has the least blurring (approximately 0.12 mm) and is capable of  producing images of calcifications in the range of 0.1-0.2 mm).

Summary

There are three specific effects of blurring in medical imaging:

• Reduced visibility of detail

• Image unsharpness

• Reduced spatial resolution

The amount (size) of blurring in a specific imaging procedure is determined by:

• Design characteristics of the imaging equipment

• Technique and protocol operating factors

In later modules we will see that blurs has different shapes, depending on the source of the blurring.

Compare the two images, what difference do you see?

All medical images have some amount of visual noise.
Noise is an undesirable image characteristic that reduces visibility of certain objects and structures.

The image on the left has a relatively low amount of noise and is the range that would be acceptable for a clinical study.

The image on the right has a high level of noise and would not be acceptable for a clinical study.

Each column contains a series of objects ranging from high contrast (bottom) to very low contrast (top).

There are different levels (low, medium, and high) of noise in the three columns.

Notice that the effect of the noise is to reduce the visibility of the low contrast objects.

Both blur and noise are generally undesirable image characteristics that reduce visibility of specific objects.

This illustration is a Contrast-Detail Diagram. Notice that the objects are arranged according to decreasing size (detail) from left to right, and according to decreased contrast from bottom to top.

The larger and high contrast objects in the lower left region should be visible under most imaging conditions.

Think of noise and blur as two characteristics that together produce a "curtain of invisibility".

Noise reduces visibility of low contrast objects.

Blur reduces visibility of small objects.

Typically, most small anatomical objects also have relatively low contrast and their visibility is reduced by both noise and blurring.

The most significant source of noise in x-ray imaging (including CT) and radionuclide imaging (including SPECT and PET) is the random statistical nature of the photons.

Noise is reduced by imaging with a greater concentration of photons which generally requires a higher radiation dose to the patient.

The noise level is also affected by the selected values of some of the imaging protocol factors as illustrated here for CT.

Adjusting an imaging procedure to reduce noise usually results in an increased exposure to the patient or an adverse effect on one of the other image characteristics, such as blurring.

The factors affecting noise will be described in other modules.

Most medical images provide one of three possible views of the human body:

• Projection

• Tomographic

• 3 Dimensional

Each view has both advantages and limitations.

Most of the tomographic imaging methods, CT, SPECT, PET, and MRI (illustrated here) have some common spatial characteristics.

Typically an anatomical region is divided into slices.

The slices are divided into a matrix (array) of tissue voxels (volume elements).

The digital image of the slice is formed as a matrix of pixels (picture elements).

The brightness or color displayed in each pixel represents some physical characteristic of the tissue in the voxel.

Images do not always display the true spatial or geometric characteristics of a body section.

Characteristics of anatomical structures and objects that can be distorted include:

• Relative size

• Shape

• Position within the body

In radiography most distortion comes from the variation in magnification with the different object locations in the body and the orientation of objects with respect to the direction of the x-ray beam.

The relative size and position of objects is distorted by the projection imaging method as used in radiography and fluoroscopy.

The magnification of an object is determined by the ratio of distances as shown.

A long (72 in) FRD is typically used for chest imaging to minimize magnification and the related distortion.

The general quality of an image is determined by the combination of the five fundamental characteristics.

Each characteristic can have an effect on the visibility of structures and objects within the body.

This is the end of this module.

To return to the beginning, CLICK HERE.