Radiation units: gray and sievert

Clinical relevance (anaesthesia)

• Most perioperative radiation exposure is from diagnostic imaging (fluoroscopy, CT, nuclear medicine) and is reported as effective dose (Sv); tissue injury risk relates more to absorbed dose (Gy) at the skin/organ.
  • Examples: C-arm fluoroscopy in orthopaedics/vascular, interventional radiology, cardiac cath lab, CT trauma imaging.
• Deterministic effects (threshold; severity increases with dose) are linked to high local absorbed dose (Gy).
  • Skin erythema, epilation, cataract, tissue necrosis; foetal malformations (timing-dependent).
• Stochastic effects (no threshold; probability increases with dose) are linked to effective dose (Sv).
  • Cancer induction and heritable effects (the latter very low/uncertain in humans at diagnostic doses).
• Staff monitoring uses personal dosimeters; limits are expressed in Sv (usually mSv).
  • Typical workplace aim: keep exposures ALARP; pregnant staff have tighter constraints.

Core definitions and relationships

• Absorbed dose (Gray, Gy) = energy absorbed per unit mass.
  • 1 Gy = 1 joule per kilogram (1 J·kg⁻¹).
  • Purely physical quantity; does not account for radiation type or tissue sensitivity.
• Equivalent dose (Sievert, Sv) = absorbed dose weighted for radiation type (biological effectiveness).
  • H_T = D_T,R × w_R
  • Units: Sv (because w_R is dimensionless).
• Effective dose (Sievert, Sv) = sum of equivalent doses to tissues weighted for tissue radiosensitivity (whole-body detriment).
  • E = Σ w_T × H_T
  • Used to compare risk across different investigations and non-uniform exposures; not intended for individual patient risk prediction.
• Key point: Gy measures energy deposition; Sv estimates biological harm/risk (via weighting factors).

Weighting factors (what you need to know)

• Radiation weighting factor (w_R) reflects relative biological effectiveness (RBE) for stochastic effects.
  • Photons (X-rays, γ) and electrons/β: w_R = 1.
  • Protons: commonly w_R ≈ 2 (varies with energy).
  • Alpha particles: w_R = 20.
  • Neutrons: energy-dependent (often quoted range ~5–20; some tables peak higher). In exams, state energy-dependent and higher than photons.
• Tissue weighting factor (w_T) reflects contribution of that tissue to overall stochastic detriment.
  • High w_T tissues classically include: bone marrow, colon, lung, stomach, breast, gonads (exact values depend on ICRP version).
  • Sum of w_T across all tissues = 1.

Conversions and typical magnitudes

• If radiation is X-ray/γ (w_R = 1): 1 Gy to a tissue ≈ 1 Sv equivalent dose to that tissue.
  • But effective dose depends on which tissues were irradiated and their w_T; it is usually much lower than local absorbed dose for a small field.
• Historical units: rad and rem.
  • 1 rad = 0.01 Gy; 1 Gy = 100 rad.
  • 1 rem = 0.01 Sv; 1 Sv = 100 rem.
• Background radiation (UK): order of a few mSv per year (varies with geography and medical exposure).

Dose quantities you may see in imaging (and how they relate)

• CTDIvol (mGy): scanner output index for CT (dose in a standard phantom; not patient-specific absorbed dose).
• DLP (mGy·cm): CT dose-length product; can be converted to effective dose using region-specific conversion factors (k).
  • E (mSv) ≈ k × DLP (mGy·cm), where k depends on body region and patient size/age.
• DAP/KAP (Gy·cm²): dose-area product in fluoroscopy; correlates with stochastic risk and is used for audit/trigger levels.
• Entrance skin dose / peak skin dose (Gy): relevant to deterministic skin injury in prolonged fluoroscopy.

Regulatory/occupational context (high-yield)

• Occupational limits are expressed as effective dose (mSv) and equivalent dose (mSv) to specific tissues (e.g., lens of eye, skin, extremities).
  • Know the concept: whole-body effective dose limit vs organ-specific equivalent dose limits; exact numbers can change with regulations—quote principles and local policy if unsure.
• Pregnancy: foetal dose should be kept very low; monitoring and work modification may be required depending on exposure environment.