Evaluation of the diagnostic performance of self-navigated whole-heart contrast-enhanced coronary MRA at 3T

Background

Recently, a self-navigated whole-heart coronary MRA technique has been developed to address these limitations of conventional coronary MRA techniques. Our study was to evaluate the diagnostic performance of self-navigated whole-heart coronary magnetic resonance angiography (CMRA) at 3T, using conventional invasive coronary angiography (ICA) as the reference.

Methods

60 consecutive patients underwent CMRA, 39 of which later underwent ICA. CMRA was performed on a 3T clinical scanner during free-breathing using an ECG-gated, fat-saturated, inversion-recovery prepared spoiled gradient-echo sequence with 3D radial k-space trajectory, self-navigated motion correction, and offline non-Cartesian sensitivity encoding reconstruction. The CMRA images were evaluated by two experienced readers to detect significant luminal narrowing (>50% diameter reduction).

Results

All patients completed CMRA successfully, with one excluded from analysis (1.6%, due to poor image quality). From the 59 included patients, a total of 506 coronary segments were evaluated. In addition, 39 of the 59 patients underwent ICA, where 315 of 367 (85.8%) segments with a reference luminal diameter ≥1.5 mm were assessable on CMRA. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy on a per-patient basis were 81.8%, 81.3%, 85.7%, 76.5%, and 81.6%, respectively.

Conclusions

Contrast-enhanced self-navigated CMRA on 3T is a promising technique for the noninvasive detection of significant coronary stenosis, and future technical improvement efforts are warranted to make it clinically viable.

Figure 1

coronary MRA multi-planar reformats with significant stenosis.

Figure 2

Representative multi-planar reformats of the coronary arteries: (a) normal RCA in patient with suspected CAD; proximal and middle segments were clearly seen; (b) left main, proximal and middle segment of LAD.