The Usefulness of Enhanced 3D-TOF MR Angiography in the Patients with Cerebral Infarction: Comparison with Conventional Angiography

Jang, Nam Kyu; Seo, Jeong Jin; Chung, Tae Woong; Jeong, Gwang Woo; Kim, Jae Kyu; Kang, Heoung Keun; Cho, Ki Hyun

Abstract

PURPOSE: The aim of this study was to compare the usefulness of enhanced 3D-TOF MR angiography with that of the conventional kind in patients with cerebral ischemic symptoms and to determine the difference between radiologists who have interpreted MR angiograms for less than one year and for more than five years. MATERIALS AND METHODS: Seventy-three patients with clinical symptoms of cerebral ischemic infarction who had undergone conventional angiography MR imaging and MR angiography were involved in this study. On the basis of divisions of the internal carotid artery, three groups were designated: Group I, from the bifurcation of the common carotid artery to the bifurcation of the internal carotid; Group II, from the bifurcation of the internal carotid to the bifurcation of the anterior and middle cerebral artery; Group III, the anterior and middle cerebral artery segments distal to their branching. Two radiologists, one who had interpreted MR angiographic findings for less than one year, and the other for more than 5 years, retrospectively reviewed the findings and graded them according to the degree of vascular stenosis. k statistics were used to measure agreement between the two readers and to compare their techniques. Sensitivity and specificity were calculated only if there were abnormal vascular findings. RESULTS: A total of 438 arteries, 146 in each group, were available. In Group I, agreement between CA and MRA was high; k was 0.538 in reader A and 0.687 in reader B and there was close agreement between the readers(K=0.621). For reader A, sensitivity was 82.4% and specificity was 77.7%, while for reader B, the figures were 88.2% and 87.2%, respectively. In Group II, agreement between CA and MRA was high; k was 0.508 for reader A and 0.566 for reader B and again there was close agreement between the two readers(k=0.622). Reader A showed a sensitivity of 88.2% and a specificity of 73.7%, while for reader B, the corresponding figures were 68.2% and 81.8%. In Group III, agreement between CA and MRA was high; k was 0.508 in reader A and 0.566 in reader B and there was close agreement between (k=0.622). For reader A, sensitivity was 50.0% and specificity was 77.6%, while for reader B, the corresponding figures were 40% and 89.7%. Overall, in total of 438 ar-teries, there was good agreement between each reader (k=0.662). Reader A showed a sensitivity of 81.1% and a specificity of 76.4%, and for reader B, the figures were 83.2% and 86.4%, respectively. CONCLUSIONS: For the evaluation of intracranial vascular disease, e3D-TOF MRA is faster and less invasive than conventional angiography. Regardless of the reader's experience, it shows high sensitivity and there is close agreement between the readers involved. It is thus a useful method for the evaluation of steno-occlusive lesions in patients with cerebral infarction.