CT

COMPUTED TOMOGRAPHY (CT)

6.1 INTRODUCTION

Core Concept

Computed Tomography (CT) uses X-rays and computer processing to produce cross-sectional images.

Rationale

A single radiograph compresses 3D anatomy into a 2D image. CT avoids this by reconstructing individual slices, allowing precise localization of structures and pathology.

Key Principle

CT images are based on differences in attenuation.

Rationale

Different tissues absorb X-rays differently due to density and atomic number. This variation is what the CT detects and converts into image contrast.

Clinical Importance

• High image quality

• High radiation dose

Rationale

Multiple projections from many angles increase diagnostic accuracy but also significantly increase radiation exposure compared to plain radiography.

Fundamental Rule

Balance image quality vs radiation dose (ALARA principle)

Rationale

Excess dose provides no additional diagnostic value once adequate image quality is achieved.

6.2 HISTORY OF CT

Key Contributors

• Radon → mathematical foundation

• Cormack → reconstruction theory

• Hounsfield → first CT scanner

Rationale

CT relies on both mathematics (Radon transform) and engineering (scanner design). Boards often test distinction between theory (Cormack) and invention (Hounsfield).

Important Dates

• 1971 → First clinical scan

• 1972 → Practical CT

• 1979 → Nobel Prize

Rationale

These milestones mark transition from theory to clinical application.

6.3 OPERATING STEPS

Sequence

1. Patient positioned on table

2. Table moves into gantry

3. X-ray tube rotates

4. X-rays pass through patient

5. Detectors capture transmitted photons

6. Computer reconstructs image

Rationale

Each step is essential:

• Tube produces beam

• Patient attenuates beam

• Detectors measure attenuation

• Computer converts data into image

Gantry Contents

• X-ray tube

• Detectors

• Rotating mechanism

Rationale

The gantry houses all imaging components required for acquisition.

6.4 GENERATIONS OF CT

Fundamental Pattern

• 1st & 2nd: Translate–Rotate

• 3rd & 4th: Rotate only

Rationale

Early scanners required linear motion for data acquisition. Later designs improved speed by eliminating translation.

Third Generation (Most Important)

• Tube and detectors rotate together

• Fan beam geometry

• Uses slip rings

Rationale

Continuous rotation (via slip rings) allows faster scanning and is the basis of modern CT systems.

Fourth Generation

• Tube rotates

• Detectors stationary (full ring)

Rationale

Detector ring allows continuous measurement but increases cost and scatter.

Fifth Generation (Electron Beam CT)

• No mechanical motion

• Extremely fast (cardiac imaging)

Rationale

Electron beam sweeps target to generate X-rays rapidly, minimizing motion artifacts in the heart.

6.5 BASIC PRINCIPLES OF CT

Attenuation Concept

X-ray intensity decreases as it passes through matter.

Rationale

Absorption and scatter remove photons from the beam, which is measured by detectors and used to calculate tissue properties.

Image Formation

• Image composed of pixels

• Each pixel represents a voxel

Rationale

A pixel is 2D, but it reflects a 3D volume (voxel) within the patient. The value assigned is an average attenuation.

Reconstruction

Filtered Back Projection (FBP)

Rationale

Simple backprojection produces blur. Filtering corrects this, resulting in a sharper image.

6.6 CT IMAGE AND HOUNSFIELD UNITS

Hounsfield Unit (HU)

CT number is based on comparison to water.

Rationale

Water is used as a reference because it is abundant in the body and provides a stable baseline.

Standard Values

• Air = –1000

• Water = 0

• Soft tissue = ~20–60

• Bone = +250 to +1000+

Rationale

These values reflect relative attenuation:

• Air absorbs almost nothing

• Bone absorbs strongly

Grayscale Display

• White = high attenuation

• Black = low attenuation

Rationale

Human vision cannot interpret thousands of HU values, so they are compressed into grayscale using windowing.

6.7 HELICAL CT

Conventional CT

• Step-and-shoot

• Intermittent scanning

Rationale

Time is lost between slices due to table movement, reducing efficiency.

Helical CT

• Continuous rotation

• Continuous table movement

Rationale

Eliminates idle time, improving scan speed and efficiency (near 100% duty cycle).

Key Technology

Slip rings

Rationale

Allow continuous electrical connection during rotation, eliminating cable limitations.

6.8 SPATIAL RESOLUTION

Definition

Ability to distinguish small objects that are close together.

Rationale

Higher spatial resolution allows detection of fine anatomical details.

Major Factors

1. Matrix Size

• Larger matrix → smaller pixels → better resolution

Rationale

More pixels divide the image into finer detail units.

2. Pixel Size

• Smaller pixel → less averaging → better detail

Rationale

Each pixel represents a smaller area, reducing loss of detail.

3. Field of View (FOV)

• Smaller FOV → smaller pixel size

Rationale

Same matrix over smaller area results in finer resolution.

4. Voxel Size

• Voxel = pixel + slice thickness

Rationale

Represents actual volume of tissue being averaged.

5. Slice Thickness (Most Important for Volume Averaging)

• Thin slice → better resolution

• Thick slice → more averaging

Rationale

Thicker slices include more tissues in one voxel, reducing accuracy.

6. Focal Spot Size

• Smaller focal spot → sharper image

Rationale

Reduces geometric blur at the source.

7. Blur

Includes:

• Motion blur

• Detector blur

• Geometric blur

Rationale

Blur reduces edge definition, directly lowering spatial resolution.

HIGH-YIELD SUMMARY

Best Spatial Resolution

• Small pixels

• Thin slices

• Small FOV

• Large matrix

Rationale

All reduce averaging and improve detail discrimination.

Worst Spatial Resolution

• Thick slices

• Large pixels

• Motion

Rationale

All increase averaging or blur, degrading image clarity.

6.9 NYQUIST SAMPLING THEOREM

Core Concept

To accurately detect an object:

Simple Interpretation

• Large pixel → object may be missed or averaged

• Small pixel → higher chance of detection

Rationale

Because objects do not align perfectly with pixels:

• They may fall within one pixel (best case)

• Across two pixels (moderate)

• Across four pixels (worst case)

This creates partial volume effect, degrading resolution.

Reducing pixel size minimizes this problem.

Board Principle

Board Trap

• Nyquist is NOT about dose

• It is about sampling and resolution

6.10 LOW-CONTRAST RESOLUTION

Definition

Ability to distinguish objects with small differences in density

Key Idea

Rationale

• Conventional radiography requires ~10% density difference

• CT can detect as low as 0.1% difference

This is why CT is excellent for:

• Liver lesions

• Soft tissue pathology

Key Factors Affecting Visibility

1. Subject Contrast

Difference between object and background

Rationale:
Greater difference = easier detection

2. Image Noise

Grainy appearance

Rationale:
Noise masks subtle differences, making small lesions harder to see

3. Window Settings

• Narrow window → better contrast

Rationale:
Compresses grayscale range, enhancing small differences

Clinical Tip

Contrast agents increase low-contrast resolution.

6.11 FACTORS AFFECTING LOW-CONTRAST RESOLUTION

1. Contrast Scale (Windowing)

• Narrow window width improves contrast

Rationale

Limits HU range → enhances visibility of subtle differences

2. Contrast-Detail Relationship

• Smaller objects = harder to see

Rationale

Even if contrast exists, small size reduces detectability

3. ROC (Receiver Operating Characteristic)

• Different observers interpret images differently

Rationale

Human perception varies → introduces subjectivity

4. Quantum Noise

• Caused by insufficient photons

Rationale

Low photon count → statistical fluctuation → noisy image

Key Relationship

Board Trap

• Increasing mAs reduces noise (not increases)

6.12 BASIC CT COMPONENTS

6.12.1 GANTRY

Components

• X-ray tube

• Detectors

• DAS

• Collimators

• Slip rings

Features

• Angulation up to ~30°

• Aperture: ~50–85 cm

Rationale

Allows alignment of anatomy and houses all major imaging components.

6.12.2 X-RAY TUBE, COLLIMATION, FILTRATION

X-ray Tube

• High mA, high kVp

• High heat capacity (3.5–5 MHU)

Rationale

Continuous scanning generates extreme heat; tube must withstand it.

Focal Spot

• Small → better resolution

• Large → better heat tolerance

Collimation

Types:

• Pre-patient

• Post-patient

Rationale

• Defines slice thickness

• Reduces scatter

• Improves image quality

Filtration

• Removes low-energy photons

Special Filter:

• Bow-tie filter

Rationale

• Reduces patient dose

• Produces uniform beam

• Improves CT number accuracy

6.12.3 DETECTOR ARRAY

Function

Converts X-ray photons → electrical signal

Types

1. Scintillation Detectors

• 99–100% efficiency

Rationale

Convert photons → light → electrical signal (very efficient)

2. Xenon Gas Detectors

• 60–90% efficiency

Rationale

Some photons lost → lower efficiency but less scatter detection

Key Concept

• Ray → single beam path

• Projection → collection of rays

Board Trap

• More photons = better signal quality

6.12.4 DATA ACQUISITION SYSTEM (DAS)

Components

• Amplifier

• Analog-to-digital converter (ADC)

Function

Converts analog signal → digital data

Rationale

Computers process only digital signals; conversion is mandatory for reconstruction.

6.12.5 PATIENT TABLE

Function

Moves patient into scanner

Key Term

• Indexing = table movement

Rationale

Precise movement ensures accurate slice positioning and helical scanning.

6.13 FAN BEAM

Definition

Diverging X-ray beam used in modern CT

Rationale

Covers entire object in one rotation, improving speed and efficiency.

6.14 FOCUSED SEPTA

Definition

Metal plates between detectors

Function

• Allow primary beam

• Block scatter

Rationale

Improves image quality by reducing scatter radiation.

6.15 IMAGE RECONSTRUCTION

Definition

Mathematical process converting projection data into image

Core Idea

Uses Radon Transform and inverse reconstruction

Rationale

Raw data alone has no meaning until mathematically processed into image form.

6.16 RECONSTRUCTION METHODS

1. Analytical (FBP)

• Fast

• Widely used

2. Iterative Reconstruction

• More accurate

• Slower

Rationale

Iterative methods repeatedly refine the image but require heavy computation.

6.17 FILTERED BACK PROJECTION (FBP)

Process

1. Collect projections

2. Apply filter

3. Backproject data

Rationale

• Backprojection alone → blurred image

• Filtering removes blur → sharp image

Board Concept

6.18 IMAGE GENERATION

1. Acquisition

• Multiple projections at different angles

Rationale

More angles = more accurate reconstruction

2. Display

• Backprojection + filtering

3. Windowing

• Adjust grayscale display

Rationale

Human eye cannot distinguish full HU range

4. Volume Visualization

• 3D reconstruction

Rationale

Improves interpretation and surgical planning

6.19 IMAGE QUALITY CHARACTERISTICS

Main Factors

• Contrast sensitivity

• Spatial resolution

• Noise

• Artifacts

Rationale

Image quality is a combination—not a single parameter.

6.20 CONTRAST SENSITIVITY

Definition

Ability to detect small density differences

Rationale

CT is superior because it detects subtle differences in soft tissue.

Key Point

• High contrast sensitivity = better soft tissue imaging

6.21 VISIBILITY OF DETAIL

Affected By

• Blur

• Voxel size

• Filters

Rationale

Blur reduces edge sharpness → decreases detectability of small structures

Trade-Off

• Improve detail → increases noise

6.22 VISUAL NOISE

Definition

Random variation in pixel values (grainy image)

Causes

• Low photon count

Control Methods

Rationale

Noise is statistical; more photons = more stable signal.

FINAL MASTER SUMMARY

Spatial Resolution

• Controlled by pixel, voxel, slice thickness

Low-Contrast Resolution

• Controlled by noise, contrast, windowing

Noise

• Controlled by dose and voxel size

Golden Rule

MINI BOARD EXAM (WITH RATIONALE)

1. Most commonly used CT generation today?

A. 1st
B. 2nd
C. 3rd
D. 4th

Answer: C

Rationale

Third-generation CT offers optimal balance of speed, efficiency, and image quality and is the standard in modern scanners.

2. Hounsfield unit of water?

A. –1000
B. –100
C. 0
D. +1000

Answer: C

Rationale

Water is the reference standard; all CT numbers are calculated relative to it.

3. Main factor affecting volume averaging?

A. Pixel size
B. Matrix size
C. Slice thickness
D. Focal spot

Answer: C

Rationale

Slice thickness determines the depth of the voxel, which has the greatest impact on how much tissue is averaged.

4. Advantage of helical CT?

A. Lower radiation
B. Faster scan
C. Fixed detectors
D. Smaller pixels

Answer: B

Rationale

Continuous scanning eliminates delays, significantly improving acquisition speed.

5. A voxel represents?

A. 2D area
B. 3D volume
C. Beam intensity
D. Detector signal

Answer: B

Rationale

A voxel includes width, height, and depth (slice thickness), representing a volume element.

1. Nyquist theorem states:

A. Pixel equals object size
B. Pixel half object size
C. Pixel double object size
D. Pixel irrelevant

Answer: B

Rationale

Sampling must be sufficient to accurately represent the object.

2. Best modality for low-contrast resolution?

A. Radiography
B. CT
C. Fluoroscopy
D. MRI

Answer: B

Rationale

CT detects very small density differences (~0.1%).

3. Main cause of quantum noise?

A. High kVp
B. Low photons
C. Large voxel
D. Filtering

Answer: B

Rationale

Noise arises from insufficient photon statistics.

4. Function of DAS?

A. Image display
B. Beam production
C. Analog to digital conversion
D. Patient positioning

Answer: C

Rationale

DAS converts detector signal into digital data.

5. Purpose of bow-tie filter?

A. Increase noise
B. Remove low-energy photons
C. Increase scatter
D. Focus beam

Answer: B

Rationale

Removes low-energy photons → reduces dose and improves beam quality.