Hounsfield Units (HU) describe X-ray attenuation relative to water: water is 0 HU and air approximately −1000 HU. They help characterize a finding, but do not identify its cause. The following are approximate teaching values for non-contrast head CT, not diagnostic cutoffs:
- Cerebrospinal Fluid (CSF): 0 to +20 HU
- White Matter: +20 to +30 HU
- Gray Matter: +35 to +45 HU
- Acute Blood (Hemorrhage): +40 to +70 HU
- Cortical Bone: approximately +1800 HU
- Calcifications: Typically >100 HU
Measured HU can vary with scanner calibration, X-ray energy, reconstruction, image noise and partial-volume averaging. Blood attenuation also changes with hematocrit and time after bleeding. Iodinated contrast raises attenuation; check the acquisition phase before using an unenhanced reference value. Changing the window changes the display, not the underlying HU.
Measure before interpreting.
On the diagnostic images, place a region of interest (ROI) within the structure, avoiding its edge, adjacent tissue and visible artifact. Review the mean and spread of values and check adjacent slices. A tiny or heterogeneous finding may not support a reliable single number. Use comparable acquisition and ROI placement when assessing change. These are measurement principles, not a substitute for lesion-specific radiology protocols.
Do not turn a number into a diagnosis.
On non-contrast CT, attenuation above 100 HU favors calcification, but blood and calcium overlap, particularly below 100 HU. HU alone cannot diagnose or exclude hemorrhage. Interpret hyperattenuation with its location, morphology, clinical context and prior imaging; radiology review and, when appropriate, dual-energy CT can help resolve uncertainty.
Elsewhere, markedly negative attenuation can support macroscopic fat, and attenuation near water can support fluid. Neither establishes a specific diagnosis; small lesions and contrast-related artifacts can mislead. Urinary stone composition also cannot be assigned reliably from a universal single-energy HU cutoff.
References: HU physics and windowing; UNC head CT teaching; cranial CT attenuation; blood–calcium overlap; clinical dual-energy CT study; reconstruction effects; ROI measurement and fluid pitfalls; stone-composition study.
