Objective To investigate the accuracy of preoperative high-resolution magnetic resonance imaging (MRI) scans to predict tumor stage, lymph node stage, and circumferential resection margin (CRM) involvement. Methods Between September 2006 and May 2009, 42 patients with histologically proven rectal cancer by the colonoscopic biopsy in Peking Union Medical College Hospital were staged preoperatively using MRI. All of the patients underwent total mesorectum excision (TME) operation within 1 week after MRI examination. The specimens were reported according to the 2002 TNM staging system for primary colorectal cancer of the American Joint Committee on Cancer (AJCC). Concordance between radiologic staging of tumor, local lymph node, and CRM involvement and pathologic reporting was assessed by means of the Kappa statistic.Results For all of 42 patients, MRI correctly staged the tumor in 36 patients, understaged in 3 patients and overstaged in 3 patients. Statistically, there was a better correlation between pathologic and radiologic tumor staging (Kappa=0.731, P=0.000). MRI correctly staged lymph node status in 31 patients, understaged in 5 patients and overstaged in 6 patients. Statistically, there was a common correlation between pathologic and radiologic lymph node staging (Kappa=0.410, P=0.009). MRI correctly reported the status of the CRM in 40 patients. Statistically, there was the best correlation between pathologic and radiologic reporting of CRM involvement (Kappa=0.829, P=0.000). Conclusion Preoperative highresolution MRI scans has a good concordance with pathologic tumor stage but common with pathologic lymph node stage. Preoperative highresolution MRI can provide reliable information about CRM and thus help to choose which patient could benefit from the preoperative neoadjuvant therapy.
In recent years, high-resolution magnetic resonance imaging (HRMRI) has become a useful clinical and research tool. HRMRI can be used to observe intracranial vascular wall lesions in vivo, providing more valuable pathophysiological information, and providing guidance for the diagnosis, differential diagnosis and prognosis of intracranial atherosclerosis. For stenotic intracranial atherosclerosis, the morphology of the vessel wall can effectively differentiate various vascular stenosis diseases. Further, plaque composition, vessel wall enhancement, remodel mode provide information of plaque vulnerability. For non-stenotic intracranial atherosclerosis, the location of the plaque can reveal the pathophysiological mechanism. In addition, HRMRI can show the lesion in lenticulostriate artery. Therefore, this article will summarize the clinical application of HRMRI.