9.2 CT Number, QA/QC & Accreditation

Why This Topic Matters

The ARRT content specifications list CT number (Hounsfield units), linearity, x-ray tube warm-up procedures, quality assurance tests (QA/QC), and accreditation as five separate sub-items (2.B.5-2.B.9) under Image Quality. These are not abstract physics trivia — they are the daily-operations backbone of a working CT department. Every image a technologist produces is quantitatively anchored to the Hounsfield scale, every scanner must pass routine QC before it can be trusted with patients, and in the United States, accreditation is directly tied to whether a facility can bill Medicare for the study at all.

CT Number (Hounsfield Units)

The CT number, expressed in Hounsfield units (HU), quantifies how much a tissue attenuates the x-ray beam relative to water. It is calculated as:

HU = 1000 × (μ_tissue − μ_water) / μ_water

where μ (mu) is the linear attenuation coefficient of the material. Two values are fixed by definition and used as universal calibration anchors: water = 0 HU and air = -1000 HU. Every other tissue falls somewhere on that scale:

Standard reconstruction historically spans roughly -1024 to +3071 HU (a 12-bit, 4096-level range), though many modern scanners support an extended CT scale (values up to roughly +32,767 HU) specifically to accommodate dense metal implants without truncating the data. This HU scale is the foundation for the windowing techniques covered elsewhere in this chapter — window level is simply the HU value centered in the displayed grayscale range.

Linearity

Linearity verifies that CT numbers scale proportionally with the true attenuation coefficient across the full clinically relevant density range — not just at one reference point like water. It is tested with a dedicated linearity phantom containing multiple inserts of known, different densities (for example, air, water-equivalent, and bone-equivalent plugs). The measured HU for each insert is plotted against its known physical density or attenuation; the result should form a straight line. A curved or offset line indicates the scanner's calibration has drifted and needs correction before it can be trusted for quantitative tasks like calcium scoring or bone density measurement. Linearity is checked on a periodic schedule (for example, quarterly or per the accreditation body's required interval) — distinct from the water-phantom check performed every single day.

X-ray Tube Warm-Up Procedures

The CT x-ray tube's anode is a rotating, high-mass component that must reach an even thermal state gradually. Tube warm-up is a manufacturer-specified sequence of exposures — typically starting at low technique and stepping up — performed:

• Before the first patient scan of the day.
• After the tube has sat idle for an extended period (commonly more than 1-2 hours, per manufacturer specification).
• After tube replacement or major service.

The purpose is to prevent thermal shock: if a cold anode instantly receives a full clinical exposure, the rapid, uneven heating can crack or damage the anode target — a component that costs tens of thousands of dollars to replace and takes the scanner out of service for days. Skipping warm-up is both an equipment-protection failure and a potential first-patient image-quality problem.

Quality Assurance / Quality Control (QA/QC) Program

A CT QC program has two tiers:

• Daily (technologist-performed): scan a water-equivalent phantom to confirm CT number accuracy for water (typically required to read 0 HU within a narrow tolerance, such as ±5 HU), acceptable noise level, and image uniformity; perform tube warm-up.
• Periodic (qualified medical physicist-performed): spatial (high-contrast) resolution using a resolution test pattern, low-contrast detectability, slice-thickness accuracy (comparing nominal versus actual slice width), CTDIvol (CT dose index) output verification against the console display, laser light/alignment accuracy, table incrementation accuracy, and artifact evaluation.

If daily QC fails — for example, the water phantom's CT number reads outside tolerance — the correct action is to remove the scanner from clinical service and notify the physicist or service engineer, not to keep scanning patients on the assumption that "it's probably fine." An out-of-tolerance daily QC result is a systemic accuracy problem affecting every study, not a one-off bad image.

Accreditation

Accreditation is a formal, external review of a CT facility's equipment performance, personnel qualifications, QC documentation, dose indices, and clinical image quality. The two major U.S. accrediting bodies are the American College of Radiology (ACR) CT Accreditation Program and the Intersocietal Accreditation Commission (IAC). A typical accreditation review requires submission of clinical images, phantom images (scanned on a standardized accreditation phantom), and a qualified medical physicist's survey covering dose benchmarks and safety.

Accreditation is not merely a voluntary quality credential. Under the Medicare Improvements for Patients and Providers Act (MIPPA) of 2008, suppliers of advanced diagnostic imaging services — including CT, MRI, and nuclear medicine/PET — must be accredited by a CMS-approved accrediting body in order to bill Medicare Part B for those services in non-hospital, freestanding settings. This is why a technologist's individual ARRT certification, the department's daily/periodic QC records, and the facility's formal accreditation status are all connected: a gap in any one can jeopardize the facility's ability to be reimbursed for the studies the technologist performs.

Exam Scenario

A new CT scanner is installed, and the qualified medical physicist completes the baseline acceptance testing and accreditation phantom imaging required for ACR accreditation. Three months later, the technologist's daily water-phantom QC shows the CT number for water reading +6 HU — outside the facility's stated ±5 HU tolerance. The correct action is to pull the scanner from clinical use and contact the physicist or service engineer for recalibration before scanning additional patients, since a drifted CT-number baseline affects the diagnostic accuracy of every subsequent study, including quantitative measurements like lesion density characterization.