Resident Physics Lectures The Radiographic Image Geometry George

Resident Physics Lectures The Radiographic Image & Geometry George David, MS, FAAPM, FACR Associate Professor Department of Radiology Medical College of Georgia

Contrast �difference in density between areas on the radiograph �Contrast depends on �subject contrast �receptor contrast �scatter

* Subject Contrast �difference in x-ray intensity transmitted through various parts of subject �Depends on �thickness difference �density difference �atomic number difference �radiation quality (k. Vp, HVL) I IS IL Subject Contrast = IS / IL

* Subject Contrast & Radiation Quality �high k. Vp = lower subject contrast � long scale contrast (less difference between areas receiving varying amounts of radiation) �low k. Vp = high subject contrast � short scale contrast (more black & white; more difference between areas receiving varying amounts of radiation) �low k. Vp increases patient dose

Scatter �Reduces contrast �Produces unwanted density �Mostly a result of Compton interactions �Increases with � k. Vp � part thickness � field size � collimation reduces scatter

Image Quality �ability of image receptor to record each point of image as point on the display �Influenced by �radiographic mottle � also called noise �sharpness �resolution

Image Quality: 3 Definitions 1. Depends only on intrinsic, objective physical characteristics of imaging system �Can be measured independent of observer �Quantitative 2. Whatever observer says it is �Subjective perception of image 3. Defined by observer’s ability to achieve an acceptable level of performance for a specified task. Courtesy Ralph Schaetzing, Carestream Health

You Already Know Some Imaging Statistics

Noise & a Die �You throw the die 6 times. Is this die rigged?

Noise & a Die �You throw the die 6 million times. Is this die rigged? 1 mi n 2 ’s o i l l i m 1 1 million 1’s llio 1 mi n 4 ’s llion 2 m illio n 6 ’s 3 ’s

Raindrops �When it first starts to rain, one can see where each drop landed �After a few minutes, sidewalk looks uniformly wet

X-Ray Images Are Created One Photon at a Time Many Photons Few Photons �Credit: Sprawls. org

Quantum Mottle �Appearance � irregular density variations in mid-density areas exposed to uniform x-ray fields �Cause �random x-ray emission �statistical fluctuations in # of quanta / unit area absorbed by receptor

Noise & Money �Which photo has more stacks of money? ? Hint: One photo has one more stack than the other

Noise & Money �Which photo has more stacks of money? ? Hint: One photo has one more stack than the other

Noise & Image Quality �Cause of noise (quantum mottle) � statistical fluctuation in # of x-ray photons forming image �Ability to see high contrast objects limited by image sharpness �High noise reduces visibility of low contrast objects � most important diagnostic information here

Similar Triangle Review Focal Spot b a h Object A B H c Receptor Object a b c h ---- = --A B C H C Receptor

Magnification Defined Focal Spot size of image ----------size of object Object Image

Using Similar Triangles size of image Magnification = ----------size of object Focal Spot h H Object Image focus to image distance H Magnification = -------------- = --focus to object distance h

* Optimizing Image Quality by Minimizing Magnification focus to receptor distance H magnification = -------------------- = --focus to object distance h Focal Spot h H Object Image �Minimize object-receptor distance �Maximize focal-receptor distance

Ever-present Imaging Artifact �Occurs whenever we image 3 D object in 2 D �Work-around �Multiple views ? ?

Sharpness �Ability of receptor to define an edge �Sharpness and Contrast � unsharp edge easier to detect under conditions of high contrast � sharp edge are less visible under conditions of low contrast �One cause of unsharpness � Penumbra � Shadow caused by finite size of focal spot

Penumbra �Latin for “almost shadow” �also called edge gradient Area source focal spot �region of partial illumination �caused by finite size of focal spot �smears edges on image �zone of unsharpness called � geometric unsharpness � penumbra � edge gradient Image

Minimizing Geometric Unsharpness �minimize focal spot size �maximize source to image distance �minimize object to image distance Minimize maximize minimize

Focal Spot Size �Trade-off �heat vs. resolving power �exposure time vs. resolving power �Focal Spot Size most critical for �magnification �mammography

Sources of Unsharpness �Geometry �Motion �minimized by short exposure times �Absorption �absorber may not have sharp edges � round or oval objects

Absorption Unsharpness �Cause �gradual change in x-ray absorption across an object’s edge or boundary � thickness of absorber presented to beam changes �Effect �produces poorly defined margin of solid objects X-Ray Tube

Total Unsharpness �combination of all the above BUT not the sum �larger than largest component �largest component controls unsharpness �improvement in smaller components don’t help much

Sharpness & Resolution �Sharpness �ability of imaging system to record sharply defined margins or abrupt edges �Resolving Power (Resolution) �ability to record separate images of small objects very close together

Distortion Types Shape Distortion X-Ray Tube Image minimal distortion when object near central beam & close to receptor Relative Position Distortion X-Ray Tube Image

Motion Unsharpness �Caused by motion during exposure of �Patient �Tube �Receptor �Effect �similar to penumbra �Minimize by �immobilizing patient �short exposure times

Inverse Square Law �intensity inversely proportional to square of distance �if distance 2 X, intensity drops by 4 X �Assumptions �point source �no attenuation Intensity a 1/d 2 d

Loss of Contrast as a Result of Unsharpness �as sharpness decreases so does contrast �less sharp system blurs dark & light areas together � maximum density decreases � minimum density increases �at very high frequency image will be uniform gray

Loss of Contrast Lowest Frequency Highest Frequency Fraction of contrast reproduced decreases at increasing frequency because lines and spaces blur into one another