Evaluation of arch vessels: breath-hold versus non-breath-hold techniques.

Abstract

Editor: I read with interest the article by Dr Randoux and colleagues in the December 2003 issue of Radiology (1). The authors made three statements that I believe are incorrect. The first (page 700) states that “on the basis of published data, a non–breath-hold technique is most appropriate for the evaluation of craniocervical vessels.” In the two references cited (2,3), however, the arch vessel origins were not evaluated, and no comparison was made between breath-hold and non– breath-hold magnetic resonance (MR) imaging. We (4) and others (5) previously demonstrated a statistically significant reduction in motion-related blurring artifacts when arch vessels were evaluated with breath holding as opposed to free breathing (4,5). As a result of these two studies, we (and many other groups) routinely perform “arch to circle of Willis” contrast material–enhanced MR angiography studies during apnea. In fact, analysis of figure 5b of the article of Dr Randoux and colleagues (1) shows the typical respiratory-induced blurring seen at the origin of the vertebral artery in question. Some of these artifacts may be minimized by means of evaluation of the source images instead of maximum intensity projections. Second, the authors state that “breath holding itself can cause artifacts.” While this may be true in rare cases, these artifacts are certainly less common than those seen with free breathing. Finally, the authors state that “we did not use the breath-hold technique because of the long imaging time (32 seconds) and the relatively old age of the study population.” First, the median patient age of 62 years (range, 32–83 years) is clearly no longer considered “old,” and a 32-year-old patient could easily hold his or her breath for 32 seconds. Second, with the use of coaching, hyperventilation, and imaging at end-inspiration (with or without supplemental oxygen), almost all patients can maintain apnea for 32 seconds unless they have severe chronic obstructive pulmonary disease or congestive heart failure. Gay et al (6) showed that most outpatients who do not smoke heavily or do not have chronic obstructive pulmonary disease or congestive heart failure can achieve a single breath hold of 38 seconds. Finally, the authors used centric phase ordering for filling of k space. Maki et al (7) have shown that this technique actually minimizes breathing-related artifacts by filling the important central lines first, providing good-quality MR angiography images even if the patient cannot tolerate apnea for the entire examination. Most of this benefit was achieved with a breath-hold fraction of as little as 50% if it occurred during acquisition of central k space, while breath holding during peripheral k-space acquisition had no substantial benefit (7).