Image noise reduction algorithm for digital subtraction angiography: clinical results

Abstract

Purpose: To test the hypothesis that an image noise reduction algorithm designed for digital subtraction angiography (DSA) in interventional neuroradiology enables a reduction in the patient entrance dose by a factor of 4 while maintaining image quality.

Materials and methods: This clinical prospective study was approved by the local ethics committee, and all 20 adult patients provided informed consent. DSA was performed with the default reference DSA program, a quarter-dose DSA program with modified acquisition parameters (to reduce patient radiation dose exposure), and a real-time noise-reduction algorithm. Two consecutive biplane DSA data sets were acquired in each patient. The dose-area product (DAP) was calculated for each image and compared. A randomized, blinded, offline reading study was conducted to show noninferiority of the quarter-dose image sets. Overall, 40 samples per treatment group were necessary to acquire 80% power, which was calculated by using a one-sided α level of 2.5%.

Results: The mean DAP with the quarter-dose program was 25.3% ± 0.8 of that with the reference program. The median overall image quality scores with the reference program were 9, 13, and 12 for readers 1, 2, and 3, respectively. These scores increased slightly to 12, 15, and 12, respectively, with the quarter-dose program imaging chain.

Conclusion: In DSA, a change in technique factors combined with a real-time noise-reduction algorithm will reduce the patient entrance dose by 75%, without a loss of image quality.