1286 MRI of Dissection of the Aorta: Recognition of the Intimal Tear and Differential Flow Velocities Robert E. Dinsmore,1 Van J. Wedeen,1 Stephen W. Miller,1 Bruce R. Rosen,1 Michael Fifer,2 G. J. Vlahakes,3 Robert R. Edelrnan,1 and Thomas J. Brady1 The use of MRI for diagnosis of aortic dissection has recently been described, including identification of the true and false channels and involvement of aortic branches [1 , 2]. The present report describes an additional finding: the iden- tification of the intimal tear in the ascending aorta in a patient with a type-A dissection. In addition, the identification of true and false aortic channels was enhanced by use of a phase- shift technique for display of differential rates of flow [3]. Case Report A 55-year-old woman was admitted to the Massachusetts General Hospital with progressive aortic insufficiency 4 weeks after having acute, severe pain in the anterior chest and neck. Echocardiography showed dissection of the ascending aorta. There was a loud murmur of aortic insufficiency. A chest radiograph demonstrated a dilated ascending aorta and normal heart size. MRI was performed with a 0.6-T system (Technicare Corp., Solon, OH) using electrocardiogram-gated spin-echo images with echo time (TE) equal to 30 msec. Images of 1-cm thickness were acquired in the short and long axes with a multislice technique. Short-axis images were obtained in a conventional image plane transverse to the thorax. Long-axis images were obtained by a technique described previously [4,5]. MA images showed a type-A aortic dissection with a partition, or flap, between true and false channels from the ascending aorta to the abdominal aorta (Fig. 1). The partition could be seen either as a thin line of medium signal intensity between two areas without signal owing to flowing blood, or as a signal arising from slowly flowing blood in the false channel contrasted to the absence of signal in the true lumen. The series of contiguous short-axis images identified the intimal tear by showing direct extension of the flow void from the true lumen into the false channel on an image slice through the lower ascending aorta, while adjacent images showed the partition sepa- rating the two channels to be intact (Figs. 1 B-i D). The image data were also evaluated by a phase-display method described by Wedeen et al. [3]. This method provides an image of phase differences induced by the motion of blood through the imaging gradients. It achieves this by superimposing a first-order phase shift on the real image component during image reconstruction. This superimposed phase shift appears as a series of stripes. The stripes remain straight within stationary tissues, but moving protons such as flowing blood cause a local shift of the stripes in the direction of motion. The case presented here showed a double phase shift within the aorta corresponding to flowing blood in the true andfalse channels separated by the intimal flap (Fig. 2). Aortography confirmed the location of the tear and the extent of dissection. At surgery, an intimal tear in the midascending aorta was found. The ascending aortic segment induding the tear was resected and replaced with a woven dacron graft, and the aortic valve cusps were resuspended. Recovery was uneventful and the patient was doing well 6 months after discharge from the hospital. Discussion Definitive surgery for dissecting hematoma consists of ex- cision of the aortic segment containing the intimal tear and exclusion of the proximal end of the false channel from pulsatile aortic flow by oversewing the resected ends, usually with interposition ofa prosthetic graft. This technique is widely used [6], although recently Miller et al. [7]found no statistically significant difference in operative mortality, functional result, or late attrition when the primary intimal tear had been re- sected compared with when it had not. In their series the intimal tear was resected in most cases (78%), however, and 5-year survival for the largest subgroup (acute type A) was 54 ± 8% when the tear was resected compared with 20 ± 16% when the tear was not resected (although the numbers are small and the difference is not statistically significant). In most centers, resection ofthe intimal tear remains an essential part of surgical management [6], and preoperative angio- graphic evaluation has emphasized determining the precise location of the proximal intimal tear. Early reports suggest that MRI may be as accurate as CT Received December 26, 1985; accepted after revision February 18, 1986. M. Fifer is a clinician scientist of the American Heart Association. This work was supported in part by the Technicare Corporation. Solon, OH. I Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA 021 14. Address reprint requests to R. E. Dinsmore, Massachusetts General Hospital. 2 Department of Medicine, Cardiac Unit, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114. 3 Surgical Cardiovascular Unit, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114. AJR 146:1286-1288, June 1986 0361-803x/86/1466.-1286 © American Roentgen Ray Society