MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

An Evaluation of Automated Quality Assurance and Physics Testing of Mobile CT Units

A Rubinstein1*, M Ahmad1 , J Feng1 , R Codina1 , P Svolos1 , (1) UTHealth McGovern Medical School, Houston, TX

Presentations

(Monday, 7/15/2019) 3:45 PM - 4:15 PM

Room: Exhibit Hall | Forum 9

Purpose: To evaluate the vendor’s daily automated imaging QA and annual physics testing of non-accredited NeuroLogica CereTom systems.

Methods: We collected data (CTDIvol, beam width, and system incrementation accuracy) from annual physics surveys of a 10-year-old, a 9-year-old, and a 5-year-old system. Technologists routinely imaged a QA phantom, and vendor software automatically calculated 11 image quality parameters. We analyzed trends and anomalies in automated QA data over a 5-year period, and compared the vendor’s automated measurements to those made on an ACR phantom using a routine brain protocol.

Results: Since installation, none of the systems failed any tests on annual physics surveys. On average, displayed CTDIvol values, system incrementation, and beam width matched physicist-measured values to within 4.6% (range:0.9–14%), 1.8 mm (0–5 mm), and 2.7 mm (1–5 mm), respectively. 13% (N=376/2930) of all auto QA tests failed, and 73% (N=274/376) of failures were due to mispositioning of the phantom. Failures not due to mispositioning were within 1.4% of vendor limits and were resolved by a rescan. Each auto QA parameter was linearly correlated with time (p<0.001) on at least two of the scanners, indicating that auto QA results can be used to track image quality drift over time. We discovered evidence of an error in the software that caused the low-contrast resolution to be consistently reported as the exact value of (and never beyond) one of the pass/fail limits. ACR tests more accurately determined low-contrast resolution (CNR=1.53,1.61,1.03). The CT number of Teflon failed ACR limits on all scanners, but passed auto QA. Ring artifacts were present on at least 37 QA scans, yet all parameters passed auto QA, which delayed engineer intervention.

Conclusion: Our longitudinal analysis revealed trends in automated QA parameters, the pitfalls of an overreliance on automated QA, and the importance of independent physics evaluations of image quality.

Keywords

Not Applicable / None Entered.

Taxonomy

Not Applicable / None Entered.

Contact Email