3.1 CT Simulation, Patient Positioning & Target Localization

Principles of CT Simulation in Radiation Therapy

\nComputed Tomography (CT) simulation is the cornerstone of modern radiation oncology planning. Unlike diagnostic CT scanners designed primarily for image quality and diagnostic evaluation, a dedicated CT simulator is engineered to reproduce the physical geometry of a linear accelerator. It integrates a wide-bore gantry, a flat carbon fiber couch top, external alignment laser systems, and specialized virtual simulation software to establish reproducible treatment coordinates.

CT Simulator Hardware Specifications

\nStandard diagnostic CT units feature a gantry aperture of 70 cm, which is frequently insufficient to accommodate patient immobilization devices such as inclined breast boards, wing boards, or stereotactic frames. Radiation therapy CT simulators require a wide gantry bore measuring between 85 cm and 90 cm to prevent collisions with positioning accessories. \nThe couch top must mimic the linear accelerator treatment couch. Diagnostic scanners use a curved couch to optimize image quality, but radiation therapy requires a flat carbon fiber table top. Carbon fiber provides low attenuation (low radiation absorption) and high mechanical rigidity, minimizing deflection under patient weight (sag tolerance < 2 mm).

---

Hounsfield Units and Electron Density Conversion

\nCT simulation provides both anatomical visualization and quantitative attenuation data essential for dose calculations. Each voxel in a CT image is assigned a Hounsfield Unit (HU) value, which quantifies photon beam attenuation relative to distilled water.

$\text{HU} = 1000 \times \left( \frac{\mu_{\text{tissue}} - \mu_{\text{water}}}{\mu_{\text{water}}} \right)$ \nWhere $\mu_{\text{tissue}}$ is the linear attenuation coefficient of the material in the voxel, and $\mu_{\text{water}}$ is the linear attenuation coefficient of water.

Clinical Hounsfield Unit Scale

• Air: -1000 HU
• Lung Tissue: -700 HU to -500 HU
• Fat / Adipose: -100 HU to -50 HU
• Water: 0 HU
• Soft Tissue (Muscle, Liver): +20 HU to +80 HU
• Contrast-Enhanced Blood: +130 HU to +300 HU
• Trabecular Bone: +200 HU to +500 HU
• Dense Cortical Bone: +1000 HU to +3000 HU

Stoichiometric Calibration Curves

\nTreatment planning systems (TPS) require relative electron density (RED) to compute photon and electron dose distributions using heterogeneity correction algorithms (such as Anisotropic Analytical Algorithm [AAA] or Collapsed Cone Convolution Superposition [CCCS]). Physicists acquire monthly scans of a tissue-characterization phantom containing inserts of known densities (e.g., LN-300 lung, inner bone, cortical bone) to plot a stoichiometric HU-to-electron-density calibration curve.

 ^
2.0 |                                    / Dense Cortical Bone
    |                                  /
1.5 |                                /
    |                              / Trabecular Bone
1.0 |-------------o--------------/ Soft Tissue / Water (RED = 1.0, HU = 0)
    |           / Fat
0.5 |         / Lung
    |       /
0.0 +------o----------------------------------------> Hounsfield Units (HU)
  -1000   -500        0       +500    +1000   +1500

---

Scan Parameters: Resolution, Matrix, and Slice Thickness

\nSelecting appropriate acquisition parameters balances spatial resolution against patient radiation exposure and data file size. Image spatial resolution depends on slice thickness, reconstruction matrix size ($512 \times 512$ standard), and Display Field of View (DFOV).

Slice Thickness Selection Protocol

1. Stereotactic Radiosurgery (SRS) & Stereotactic Body Radiation Therapy (SBRT): 1.0 mm to 1.5 mm slice thickness. High spatial resolution is required to contour micro-structures (e.g., cochlea, optic chiasm) and minimize volume averaging errors in small targets.
2. Head & Neck / Thorax / Abdomen / Pelvis (Standard 3D-CRT & IMRT): 2.5 mm to 3.0 mm slice thickness. Provides an optimal compromise between spatial fidelity and reconstruction speed.
3. Palliative Large-Volume Scans (e.g., Whole Brain, Whole Femur): 3.0 mm to 5.0 mm slice thickness.

$\text{Pixel Size} = \frac{\text{DFOV (mm)}}{\text{Matrix Size (e.g., 512)}}$ \nIf a scan uses a $500\text{ mm}$ DFOV and a $512 \times 512$ matrix, the spatial pixel resolution in the transverse plane is $500 / 512 = 0.98\text{ mm}$.

---

Motion Management: 4D-CT Simulation & Respiratory Gating

\nTumors located in the lungs, esophagus, liver, and pancreas experience organ motion due to respiration, moving up to 2-3 cm during normal breathing. Conventional 3D-CT scans snapshot the tumor at an arbitrary phase of the breathing cycle, potentially causing severe target missing or artifact distortion.

4D-CT Acquisition Methodology

4D-CT tracks breathing using an infrared reflective marker block placed on the patient's abdomen (e.g., Varian RPM system) or a spirometric airflow sensor. The CT scanner operates in cine or low-pitch helical mode, acquiring images continuously at each couch position throughout at least one complete respiratory cycle (typically 3-6 seconds).

4D-CT Dataset Reconstructions

• Phase-Based Reconstructions: Images are binned into 10 respiratory phases (0% = end-inspiration, 50% = end-expiration).
• Maximum Intensity Projection (MIP): Reflects the highest Hounsfield Unit value encountered at each voxel across all 10 phases. Excellent for highlighting high-density lung nodules moving within low-density lung parenchyma to construct the Internal Target Volume (ITV).
• Average Intensity Projection (AIP): Averages the HU values across all phases, representing the mean electron density map for accurate heterogenous dose calculation in moving targets.
• Deep Inspiration Breath-Hold (DIBH): Utilized extensively in left-sided breast cancer to displace the heart inferoposteriorly away from the internal mammary and anterior chest wall treatment fields, significantly reducing mean heart dose.