Daily Quality Control: Water Calibration, Noise & Field Uniformity

Daily quality control: establish that the scanner is ready

Quality assurance (QA) is the overall program that keeps equipment performance and clinical imaging reliable. Quality control (QC) consists of specific measurements within that program. A satisfactory water phantom does not prove that every clinical protocol is optimized; it does establish a reproducible check for drift, noise and conspicuous artifacts before clinical use. The technologist performs the assigned tests, records results and escalates failures. A qualified medical physicist establishes or approves methods, baselines and action limits.

The ACR's current CT QC requirements call for daily water CT-number and standard-deviation measurements and artifact evaluation. The manufacturer may require additional startup checks. Monthly visual checks of equipment and monitors, applicable printer checks and the annual physicist evaluation are separate parts of the program. Do not substitute an annual accreditation phantom result for the daily constancy test.

Warm-up and calibration are different operations

A tube warm-up controls thermal loading after the tube has been idle. Follow the scanner's instructions for when it is needed and which automated sequence to use. There is no universal exposure sequence, idle-time cutoff or tube temperature applicable to every CT system. Warm-up should be performed with the room clear and the prescribed safety checks completed. A system designed for a different anode or bearing architecture may use a different preparation sequence.

Air calibration measures the detector response without an attenuating object so that channel sensitivities and reference signals can be corrected. It may be scheduled automatically, requested by the manufacturer or performed during troubleshooting. It is not a patient scan and should not be performed with the water phantom or an unrelated object inadvertently in the beam. A water calibration is a manufacturer-specific procedure for correcting image behavior, whereas a water phantom QC scan independently measures the resulting behavior. Passing calibration software does not replace checking the reconstructed image.

Reproduce the measurement conditions

Use the prescribed phantom, support, acquisition and reconstruction. Center the phantom in the scanner and ensure that the selected slice passes through homogeneous water rather than a cap, bubble or support structure. Record kVp, mAs or modulation mode, collimation, reconstructed thickness, kernel, field of view and reconstruction method. Changing these settings can change noise even when the hardware is functioning normally.

Place the region of interest (ROI) in homogeneous water, away from the phantom wall and obvious bubbles. Use the program's specified size and placement so that results can be compared with baseline measurements. Record the mean CT number and the standard deviation (SD), not just an impression that the image looks smooth. Evaluate the image with a window narrow enough to reveal subtle rings, streaks and nonuniformity. Review additional slices when the local procedure requires them.

Interpret limits in their proper context

Use the local, physicist-approved daily action limits. Do not copy an accreditation material tolerance into a daily constancy policy without approval. For example, the ACR accreditation phantom water interval is −7 to +7 HU under its stated testing conditions. That does not mean every institution must use the same range for its daily water test; a tighter local constancy limit may be appropriate. Similarly, a change in SD must be compared with a baseline obtained with the same technique.

Consider a hypothetical baseline of 5.0 HU SD using a particular protocol. Today's SD is 6.0 HU. The increase is (6.0 − 5.0)/5.0 × 100 = 20%. That calculation describes the change; whether it exceeds an action limit depends on the approved QC program. Do not declare a universal 20% pass or failure threshold from the arithmetic alone. If today's reconstruction is half as thick, the comparison is confounded and the original settings should be restored before interpreting hardware performance.

Respond to a failure systematically

First verify setup and technique, then repeat the test if the procedure authorizes a repeat. A persistent ring in a correctly positioned water image suggests a detector or calibration problem rather than patient motion. A mean shift with unchanged noise raises a different question from a large noise increase with a stable mean. Save representative images and record the measured values, settings, time and corrective actions.

Notify the responsible physicist, supervisor or service personnel under the program's escalation pathway. The responsible team decides whether the scanner can continue clinical work, whether particular protocols must be restricted, and what repair and retesting are required. Never conceal a failure by altering the baseline, increasing mAs or deleting an unfavorable result. After repair, verify the affected performance before returning the equipment to unrestricted use.

Daily QC is therefore a reproducibility exercise with clinical consequences: fixed inputs make a meaningful trend possible, image review finds structured errors that a single ROI can miss, and documentation preserves the evidence needed to investigate a change.

References: ACR CT quality control requirements, ACR CT accreditation phantom testing.