“Helping educational programs make digital radiography easy to understand.”

Ten Common Misconceptions in Radiography

Quinn B. Carroll, MEd, RT

 

 

 

Accuracy Vs. Simplification:

As educators (and scientists), simplification is desirable

-However:

-Caution!

Simplification often leads to some inaccuracy

Over-simplification can be misleading

 

X-ray absorption by air matters:

“The greater the SID, the more absorption by air”

Air Absorption in Perspective

One cubic meter of air (at typical pressure and    temperature) has a mass of 1.23 kilograms

A 10 cm cube of soft tissue has a mass of 1.24 kilograms, a perfect match

Therefore, one Gray delivered to a square meter of air will deliver one Gray to each 10-cm cube of tissue within the body (each cube about the thickness of the lower leg)

Body tissue is approximately 1000 times denser than air

Visualizing Volumes

To visualize one cubic meter of air, picture a cube with dimensions of the SID above the tabletop on every side 

To visualize a 10-cm cube, picture a cube about the thick-ness of the mid-lower leg  

Using an “average” collimation at the tabletop of 33 cm X 33 cm (13 X 13 in.)   

To fill this cubic meter with an average x-ray beam of 33 cm width (~13”) requires 9 x-ray beams to pass through it

For air to absorb the same amount of x-rays from a single x-ray beam, the x-ray tube would have to be 9 meters away –

This is 30-feet SID !!

(360 inches)

I.e., An SID of 30 feet (360 inches) would be required for air to absorb as much radiation as the lower leg!

Air absorption is clinically negligible

Especially compared to the effectiveness of simple geometry (inverse square law)

Less than 10% effect of geometry

The Inverse Square Law

Accounts for practically all effect of SID on beam intensity

Air absorption is negligible

It is pure projection geometry

Nothing to do with “air absorption”

Reductions to patient dose are due entirely to the inverse square law, not to air absorption

 

Increased SID to Reduce Patient Dose

Several manufacturers are now producing x-ray tables with an increased default SID set at 110 cm (Agfa), 115 cm (Siemens)

In addition to improved sharpness, reduced magnification, and increased field-of-view, increased SID also slightly reduces patient entrance skin dose (ESD)

Increased SID can save patient skin dose even after compensating the mAs for the effect of the inverse square law, because SSD (source-to-skin distance) is changed by a greater ratio than the SID:

Example:

Assume a patient of 20 cm thickness, and an increase in SID from 100 cm to 125 cm.

The SID itself has been altered by a 25% change.

Proportionately, how much was the SSD (source-to-skin distance) changed?

The original SSD was 100 – 20 = 80 cm

The new SSD is 125 – 20 = 105 cm

The original SSD was changed by 31% (25/80)

Both distances were increased by 25 cm as a raw number, but because the original SSD was less than the original SID, this change represents a greater proportional change for the SSD (31%) than for the SID (25%)

Using the square law, to compensate for the increase in SID from 100 to 125 cm, the mAs must be increased by 1.56 X

Using the same type of square law calculation for the SSD, the needed compensation for the increase in SSD would hypothetically be 1.72 X

Taking the ratio 1.72/1.56 we get a difference of 1.103 times. This is a 10.3 percent difference between the mAs compensation we actually used and the compensation that would have been required for the SSD increase

The net result is a slight reduction in patient dose (~10%) even after compensating mAs for the increased SID

Radiographers should consider using the maximum SID allowed by the equipment for tabletop exposures

Older units may lock out exposure at substantially changed SID

-BUT do not prevent us from taking advantage of 72” (180 cm) SID at upright “chest board”

 

Even if mAs is compensated, there is a slight net savings in patient dose

Entirely due to the mathematics and geometry (ISL), NOT air absorption

Note: Because of the increased exposure latitude of digital imaging, a 10-15% increase in SID can be executed without compensating mAs for even greater savings in patient dose

(provided that the combination of increased SID and other factors that reduce exposure to the IR do not add up to such an extent that unacceptable mottle becomes apparent in the image)

Using 72” (180cm) SID Can Reduce Patient Dose by 33%

-Assume patient 20 cm

-SID increase, 40” to 72”, would normally require 3X mAs

 

Original SSD was 100 – 20 = 80 cm

New SSD is 180 – 20 = 160 cm

 

The original SSD was increased by double (80 to 160 cm)

By ISL, this would require 4X mAs

 

Ratio between compensations is 4/3 = 1.33 = 33% difference

 

Greater proportional change for SSD results in 33% saving in patient dose, because mAs was increased 3X not 4X

 

Conclusion: Doing procedures upright at the chest board with 72” (180 cm) SID saves 33% patient skin dose, even when mAs is compensated

-Not recommended for Odontoid where magnification of the open mouth is an advantage

 

“A darker image has lower contrast”

Within the normal range of exposures, density can be doubled (left) without changing contrast

Extremely overexposed images (shoulder of H&D curve) or extremely underexposed images (toe of H&D curve) are the exception, not the rule, and should not be taught as the rule

Within the normal range of exposures, density can be increased (brightness reduced) without changing gray scale (contrast)

Extremely overexposed images (shoulder of H&D curve) or extremely underexposed images (toe of H&D curve) are the exception, not the rule, and should not be taught as the rule

NOTE that digital dynamic range is typically displayed with no “shoulder” or “toe” to the curve

Shoulder implies that background density cannot get any darker because all silver bromide crystals are “used up”

This does NOT apply for digital dynamic range

-Very extensive range of numbers, not limited number of crystals

 

Within the normal range of exposures, gray scale can be lengthened (lowering conrast) without changing average brightness

Extremely overexposed images (shoulder of H&D curve) or extremely underexposed images (toe of H&D curve) are the exception, not the rule, and should not be taught as the rule

To avoid contradictions,

Contrast must be defined as a difference between tissue areas (anatomy), not between the entire anatomical part and the background

E.g., between soft tissue and bone (joint space and patella)

“A darker image has lower contrast” confuses two image qualities (density and contrast) as having a cause-and-effect relationship

Causes should always be variables, NOT other image qualities

OPTIONAL: Also leads to a multitude of false conclusions, e.g.:

mAs controls contrast

SID controls contrast

 

These variables change all interactions by equal proportions

To change contrast, the RATIO between photoelectrics, Comptons, and penetrating x-rays must be changed

 

OPT: Demo: Use Ratios (not subtraction) for Contrast

Densities 3 and 1:             -Add fog at +1 …

Example:  Effect of fog on subject contrast:

Dividing:        3/1 and 4/2         -Results: 3 Vs. 2  C↓

Vs. Subtracting:          3-1 and 4-2         -Results: 2 Vs. 2  C=       ALL WRONG

 

Example:   Effect of Doubling mAs or ¾ SID on subject contrast:

Dividing:   2/1 and 4/2             -Results: 2 Vs. 2  C=

Vs. Subtracting:          2-1 and 4-2         -Results: 1 Vs. 2  C↑     ALL WRONG

 

“Umbra is the sharp portion of a detail”

Umbra Vs “Sharpness”:

Some publications have defined the umbra as the “sharp” portion of the image – This is false

The concepts of sharpness and unsharpness both refer to the width of the penumbra at the edge –

Physicists call extent of the penumbra edge spread

Unsharpness or blur is the extent of penumbra

Sharpness is the lack of penumbra

“Sharpness”

When a very wide penumbra is present, (left), the image is unsharp or blurry

With a very narrow penumbra, (right) the image is sharp

“Umbra” Vs “Sharpness”

Umbra can remain unchanged while sharpness changes

Left:

Unsharp image with 4 cm umbra

Right: Sharp image with 4 cm umbra

Therefore, “sharpness” is unrelated to the umbra

Umbra and Penumbra:

The umbra is characterized by its homogeneity

The penumbra is transitional in its density

-It is what is happening to the density or brightness that defines these two regions

 

“Scatter affects sharpness (spatial resolution)”

Image Sharpness Vs Contrast :

Scatter radiation reduces contrast at the edge of an image detail (reducing its visibility), C

Scatter is unrelated to the formation of penumbra at the edges of the image

Scatter radiation is NOT related to sharpness (spatial resolution)

Scatter Vs. Blur

Scatter cannot affect spatial resolution because it is not part of the projection geometry of penumbra formation

Below, the measured width of penumbral blur is 5 mm   regardless of whether scatter is present or not

The effects of scatter and blur are often confused – but they are separate in origin, nature, and effects:

 

Scatter:                                                Blur:

Completely random                        Geometrically predictable

Affects image visibility                    Affects image recognizability

Affects general area                        Affects only image edges

Emanates from patient  Emanates from x-ray tube (focal spot)

Scatter affects all three visibility functions in the image (exposure, subject contrast, and noise), but none of the recognizability (geometrical) functions

 

High kVp causes “more” scatter

Scatter Production:

At higher kVp levels, beam penetration increases

Higher penetration means fewer interactions of all kinds – fewer photoelectrics, fewer Comptons

Therefore, how can the image receptor receive more “fogging” ?

i.e., how do more scatter photons end up in the remnant x-ray beam?

3 Answers:

#1: Photons that are scattered have higher energy

Scatter photons with higher energy tend to be emitted in a more forward direction, so a higher percentage of produced scatter reaches the IR

#2: Photons that are scattered have higher energy

Scatter photons with higher energy penetrate out better to escape the body and reach the IR

#3: Photoelectric interactions plummet at higher kVp, while Compton     interactions only decrease slightly.

Therefore, within the remnant beam, Compton scattered photons comprise a much higher percentage of the total information reaching the IR than photoelectrics represent

OPTIONAL:  High kVp Levels Maximize Information:

Produce lengthened gray scale in the remnant beam signal to maximize information in the

Image

 

To shorten gray scale, A, the computer can sample known data (every other density).

But, to lengthen gray scale that is too short, B, the computer must interpolate between densities, effectively fabricating information.

Benefits of High kVp Technique:

  1. Ensures adequate penetration to produce subject contrast
  2. Combined with low mAs, it reduces patient exposure
  3. Produces lengthened gray scale in the remnant beam signal to maximize Information in image

 

DIGITAL IMAGING

“Garbage in, garbage out”

– Military and intelligence organizations routinely use digital enhancement capabilities on satellite images, leveraging magnification, contrast enhancement, interpolation and detail enhancement to demonstrate people and objects not discernible in the original image

-Medical applications can do the same thing
-Digital processing cannot create information that is literally not present in the original data, however …
-As regards the power of the human eye, it can bring out details that were previously invisible

-KEY: Digital software can operate on levels below the human range of vision
-spatial resolution too small to see
-contrast resolution too subtle to see

Example: Technically, analog information is more accurate than digital information, yet even though digitization “rounds out” values discarding analog information, digital clocks used at the Olympics are more accurate than the human eye watching an analog clock for readout accuracy

 

A very dark image is “saturated”

If any details at all can be made out in the dark portion of the image, it is over-processed but NOT

saturation

True saturation presents a flat black area with absolutely no details present

The reason the Deviation Index table urges “no repeat” for images that are simply too dark is because these are caused by failures in digital processing and can generally be corrected through additional windowing

True saturation, on the other hand, is an electrical phenomenon that occurs at the detector, not during digital processing. Therefore, it cannot be corrected by windowing

Like filled buckets spilling water over into each other, electrical charge can overwhelm the capacity of the dexels to store it, spilling across an area of the detector plate. This is saturation

With all read-outs at maximum value, there ceases to be any distinction between pixels, i.e., no image to process

Extreme over-exposure can overwhelm the digital detection system, causing a loss of data that results in a flat black appearance of the over- exposed portion of the image, (lungs, right)

This is not “fog”, but a loss of data

It takes at least eight to ten times the normal exposure to reach saturation

-For several manufacturers, it takes even more

 

This corresponds to an exposure index number of 3000 for the CareStream (Kodak) system, or an S number of 25 for Fuji

So, saturation is rare, and in most instances of overexposure the computer system is able to perfectly compensate such that the final image has good quality

 

“Fog” is always transmitted through to the displayed image

In the same sense that a fogged film developed into a fogged image

 

Upon digitization, all information is in the form of numbers

Although numbers can be “corrupted,” they can also be manipulated and corrected in ways chemical fog could not

Contrast Comparisons for Single-Step Increases in kVp

Each Step: 15% ↑ kVp & ½ mAs

Conventional Film Radiographs Vs. Digital CR Images

Note that CR images do not show the progression of fog apparent in the film images

Conventional film radiographs of the lateral lumbar spine (left) often showed a fog pattern from table scatter (arrows), which obliterated the posterior portion of the spinous processes

Digital processing is able to correct for this on most lumbar radiographs (right), such that the spinous processes are demonstrated entirely (arrows)

About 3 out of 4 film images suffered from the fog pattern

About 3 out of 4 digital images correct for it

Digital equipment is remarkably resilient to the effects of scatter radiation caused during an exposure

The “garbage in, garbage out” philosophy implies that computer algorithms cannot compensate for “raw” data that has been affected by scatter radiation, i.e., “garbage.” This is largely false –

The post-processing capability of digital imaging allows for correction of just such errors as excessive darkness or excessive gray scale that scatter can cause

 

OPTIONAL: Pre-Fogging of CR Plates

If a CR plate has been “fogged” by either background or scatter radiation prior to using it for a radiographic exposure, a digital image with poor contrast will generally result

Digital processing is NOT generally able to correct for the effects of pre-fogging a CR plate

DR image receptor plates, which are automatically erased between exposures, are not vulnerable to “pre-fogging” – a “fogged” appearing image is more rare for DR

“Pre-fogging” CR Cassettes Vs. Fogging During Exposure:

“Pre-fogging” a CR plate with background or scatter radiation will usually cause histogram analysis and exposure indicator errors

Scatter radiation that occurs during an exposure usually does not result in histogram analysis or exposure indicator errors

What is the difference?

Most typical localized fog patterns (bottom graph) do not alter the shape of the acquired image histogram enough to throw off landmark identification

-ADD typical FOG:

(Bottom Graph)

If SMAX has been defined as the second time the threshold is reached, scanning from right to left,  histogram analysis looking for the SMAX point within the anatomy on this histogram will still be able to locate it

However, “pre-fogging” of a CR cassette affects primarily the lighter portions of the image, adding a large number of light gray pixels at the left of the histogram, and leaving no “blank” white pixels present. This skews the location of SMIN and SAVE , and  can remain uncorrected.

(Demonstrations of the sensitivity of CR plates to scatter use “pre-fogging” events rather than “fogging” during an exposure)

 

Digital processing is very robust for eliminating “scatter effect” generally

– Two exceptions:
1. Pre-fogging of CR cassettes
2. Extreme cases of fogging
kVp still “controls” contrast

What “controls” contrast in the Digital Age?

– kVp controls subject contrast in the remnant beam signal reaching the IR

-Displayed contrast of the final image is primarily controlled by the LUTs used in gradation             processing

“Controlling Factors” for Displayed Digital Image Qualities

Since the advent of digital imaging, only one of the five image qualities (shape distortion) can still be attributed to the “controlling factors” traditionally taught

These old “controlling factors” still apply to the  latent image carried by the remnant beam to the IR, but not to the final displayed digital image

 

The EI/DI is the indicator of actual patient dose

Inappropriate Use of the
Deviation Index

The deviation index is:

An indicator, not a measurement, of patient exposure

Taken from data at the image receptor, (not at the patient’s surface or within the patient, even in any simulated sense), the DI is a “guide to exposure intended to indicate the acceptability of signal-to-noise ratio conditions”

 

The EI/DI is the indicator of image quality

Inappropriate Use of the
Deviation Index

“Even if images being produced clinically have corresponding DI’s well within the target range, the clinical techniques used may still not be appropriate. One can just as readily achieve an acceptable DI for an AP L-spine view with 65 kVp as with 85 kVp; evidence of underpenetration and concomitant excess patient exposure with lower kVp may be … windowed and leveled out in a digital image. Similarly, poor collimation, unusual patient body habitus, the presence of prosthetic devices, or the presence of gonadal shielding in the image may raise or lower DI’s (depending on the exam and projection) and perhaps hide an inappropriate technique.”      -AAPM Task Group 16, JMP, Vol. 36, No. 7, July 2009

The deviation index is:

One indicator, not the indicator, of image quality

  1. Leveling and windowing can hide inappropriate initial technique
  2. Several positioning-related factors can hide an inappropriate initial Technique

Speed class should be set by the manufacturer (and left unchanged)

Nearly all modern CR and DR systems can be operated at a speed class of 300, 350, or 400, without the appearance of substantial mottle in the image

This ability should be a primary consideration in purchasing digital equipment

Imaging managers, QC techs, and supervisors have a choice in determining speed class

Radiologists can be involved in determining the levels of mottle that can be allowed for different procedures