Digital variance angiography (DVA) is an alternative image processing method for conventional angiography with an image quality advantage over digital subtraction angiography (DSA). Our aim was to investigate whether the quality reserve of DVA is suitable for radiation dose reduction during lower extremity diagnostic angiographies. A secondary aim was to compare two DVA algorithms, DVA1 and DVA2. The former is the original version, whereas the latter was developed to further improve image quality of acquisitions with higher background noise.
This prospective study enrolled 114 peripheral artery disease patients (Fontaine IIb-IV, 72.8% male, mean age 66(IQR,9.25) years) undergoing lower extremity diagnostic angiography between April and September 2020. Patients were block-randomized into standard-dose (SD) (max.1.2-µGy/frame, n=57) or low-dose (LD) acquisition protocol (max. 0.36 µGy/frame, n=57) enhanced by DVA. Angiographic acquisitions of the aortoiliac, femoral, popliteal, and talocrural regions were performed. DSA images were generated in both groups (SD-DSA, LD-DSA). DVA images were generated only in the LD group (LD-DVA1, LD-DVA2). Total and angiography-related radiation dose area product (DAP) was compared between SD and LD groups. Image quality was assessed on a 5-grade Likert scale by six independent vascular interventional specialists. Data were analysed by the Kruskal-Wallis test followed by Dunn test, or by the Wilcoxon signed-rank test, where appropriate.
Total and angiography-related DAP (expressed in µGy*m2) was significantly lower in the LD group compared to the SD group (642.3(IQR,614.8) vs 1044.8(IQR,1417.3), p<0.001), and 279.3(IQR,268.1) vs 724(IQR,1002.6), p< 0.0001), respectively. The low-dose protocol considerably impaired the image quality of DSA images, LD-DSA scores were significantly lower than SD-DSA scores in all regions except in the popliteal region. On the other hand, the visual evaluation scores (Fig.1) did not differ significantly between SD-DSA and LD-DVA1 or LD-DVA2 images, except in the popliteal region where the LD-DVA2 score was significantly higher. Comparison of the DVA algorithms showed that there was no difference between DVA1 and DVA2 in the aortoiliac region, but DVA2 received significantly higher scores in the femoral, popliteal and talocrural regions and the advantage of DVA2 was the most emphasized in the talocrural region.
DVA imaging allowed a highly significant reduction in total (-40%) and angiography-related (-62%) radiation dose in lower extremity angiography, without affecting the image quality. Low-dose DVA2 images compared to standard dose DSA images showed a tendency of even better visual scores in the popliteal and talocrural regions, and outperformed DVA1 in the femoral, popliteal and talocrural regions, therefore this algorithm might be especially beneficial in below-the-knee imaging.