CHICAGO – Coronary computed tomography angiography with diagnostic image quality is feasible at an ultralow radiation dose of 0.2 millisievert using model-based iterative reconstruction.
This represents roughly an 80% reduction in radiation dose compared with standard coronary CT angiography, Dr. Julia Stehli said at the annual meeting of the Radiological Society of North America.
Increasing concerns about radiation exposure have prompted the use of prospective ECG triggering to reduce radiation doses from 20 mSv to 2 mSv or less. Several vendors also have developed new raw-data–based iterative reconstruction algorithms to further reduce radiation doses, but the trade-off can be increased image noise.
The model-based iterative reconstruction (MIBR) algorithm (GE Healthcare), however, has shown promising results for noise reduction, said Dr. Stehli of University Hospital Zurich. The technology, known as Veo, is already in use in the United States, Europe, and Asia for abdominal CT scans but is not yet commercially available for cardiac scans because of the added complexity of ECG triggering.
Dr. Julia Stehli
Dr. Stehli reported on the hospital’s first clinical experience with MIBR in 25 consecutive prospectively enrolled patients with suspected coronary artery disease who underwent standard low-dose coronary CT angiography (CCTA) and same-day ultralow-dose CCTA on a 64-slice CT scanner with prospective ECG triggering. Tube voltage and current were adapted to body mass index, which covered a wide range from 18.4 kg/m2 to 40.2 kg/m2. Contrast media volume and flow rate were adapted to body surface area. Intravenous beta-blockers were used prior to CCTA in 20 patients.
Standard CCTA was reconstructed using 30% of adaptive statistical iterative reconstruction (ASIR) according to usual hospital practice, while the ultralow-dose images were sent to the vendor for reconstruction with MIBR.
The effective radiation dose was 1.3 mSv with standard CCTA and 0.2 mSv in the ultralow-dose CCTA group (P less than .001), which is in the range reported for a postero-anterior and lateral chest X-ray, Dr. Stehli said.
A total of 100 vessels and 330 coronary artery segments were semiquantitatively assessed by two blinded, independent readers using a 4-point Likert scale, with 1 being nondiagnostic, 2 good, 3 adequate, and 4 excellent. The Kappa value for interobserver agreement of image quality was 0.8.
The average image quality score per segment was 3.3 with standard CCTA vs. 3.4 with ultralow-dose MBIR (P less than .05), she said.
Diagnostic image quality (score 2-4) was found in 319 segments (97%) and 317 segments (96%), respectively.
"These numbers are quite revolutionary," session comoderator Dr. Konstantin Nikolaou, professor of radiology at the University of Munich, said in an interview. "We’ve heard about 1.0 [mSv], so 0.2 [mSv] is great."
Still, more details are needed on exactly how the protocol works and the need to send images to the vendor for MBIR reconstruction, he said.
During a discussion of the results, Dr. Stehli said that reconstruction by the vendor typically took about 15 minutes, but Dr. Nikolau said that "it’s hard to say if that is feasible in routine clinical practice."
The ultimate test for the ultralow-dose protocol will be the clinical outcomes data, expected to be reported in 2014.
"If that proves to be robust and works in many patients and rates a good diagnostic accuracy, it would be great," Dr. Nikolau said.
The investigators would not release details on the clinical outcomes but said sensitivity and specificity for the new protocol are good.
"We believe this will have clinical applications in the near future," Dr. Stehli said in an interview.
Most patients in the study presented with chest pain (72%), and 56% were smokers, 44% had arterial hypertension, and 36% had a family history of cardiovascular disease. Their mean age was 58 years.
Dr. Stehli and her associates reported having no financial disclosures.