医学
放射科
活检
预测值
核磁共振扫描
病态的
计算机断层摄影术
磁共振成像
脑瘤
核医学
病理
内科学
作者
S H Taghipour Zahir,M Rezaei Sadrabadi,F Dehghani
出处
期刊:Iranian journal of pediatric hematology and oncology
[Knowledge E]
日期:2011-12-10
卷期号:1 (4): 121-125
被引量:15
摘要
Background Cerebral neoplasm arises from brain, spinal cord and meningeal cells. Not only malignant cerebral neoplasm also benign tumor could lead to death due to mass effect on vital structures. Access to these tumors is difficult, and MRI and CT scan could be helpful in determining anatomical location of tumors and distinction of malignant from benign. Objective For better and earlier diagnosis and treatment, present study determined the accuracy of MRI and CT scan in compare with pathological findings. Materials and Methods This experimental case-series study compare the results of tumor imaging (MRI and CT scan) with biopsy in patients who came with brain mass between April 2004 and April 2010. Demographic characteristic and medical history were recorded. The results of CT scan, MRI and biopsy reports were recorded for patients, and all data compared and analyzed by SPSS software version 15. Results Results of 218 patients were analyzed. 189 patients had definite diagnosis using CT scan, which 13 (7.2%) were diagnosed benign and 159 (92.4%) malignant. Sensitivity, specificity, positive predictive value and negative predictive value of CT scan in comparison with biopsy were 83%, 10%, 93% and 3% respectively. The accuracy of this method was 78%. Fifty four patients (24.4%) were evaluated by MRI. Sensitivity, specificity, positive predictive value, negative predictive value and accuracy of MRI were 92%, 25%, 93%, 2% and 87% respectively. Conclusion According to the results, positive results by MRI and CT scan are valuable and have diagnostic value, but negative reports need more evaluation and no roll out malignant tumor. So remarkable specificity, but low sensitivity were achieved for CT scan and MRI. They accepted as easier and more accessible methods to approach brain tumors.
科研通智能强力驱动
Strongly Powered by AbleSci AI