PURPOSE: To determine the accuracy of preoperative neuroradiological diagnosis of pathological intracranial processes in a prospective study. MATERIALS AND METHODS: A team of three neuroradiologists determined the diagnosis in CT, MR and angiography prior to stereotactic biopsy or operative resection (173 patients). Only one diagnosis was allowed, except in those cases with two equally probable diagnosis (24 patients). In 106 patients a resection of a brain tumor, in 67 cases a stereotactic biopsy was performed. According to the histological diagnosis the patients were subdivided into three groups: 1: complete agreement: the single diagnosis was correct. 2: conditional agreement: on of the 2 differential diagnosis was correct. 3: no agreement RESULTS: In 131 cases (76%) a complete agreement, in 24 cases (14%) a conditional agreement and in 18 patients (10%) no agreement were found. Assuming only stereotactic procedures the neuroradiological diagnosis was correct in 44 cases (66%) and incorrect in 10 cases (15%). In 13 patients (19%) one of the two differential diagnosis was correct. The specificity of the major tumors was calculated between 92% and 100%. The sensitivity for pituitary adenomas (n = 9) and neurinomas (n = 11) was 100%, the sensitivity for meningiomas (n = 32) was 94%. A sensitivity was calculated between 50% and 71% for astrocytomas (WHO I to WHO IV, n = 64) and metastases (n = 24). CONCLUSION: The accuracy was found to be higher than in the comparable retrospective studies.
PURPOSE: To determine the accuracy of preoperative neuroradiological diagnosis of pathological intracranial processes in a prospective study. MATERIALS AND METHODS: A team of three neuroradiologists determined the diagnosis in CT, MR and angiography prior to stereotactic biopsy or operative resection (173 patients). Only one diagnosis was allowed, except in those cases with two equally probable diagnosis (24 patients). In 106 patients a resection of a brain tumor, in 67 cases a stereotactic biopsy was performed. According to the histological diagnosis the patients were subdivided into three groups: 1: complete agreement: the single diagnosis was correct. 2: conditional agreement: on of the 2 differential diagnosis was correct. 3: no agreement RESULTS: In 131 cases (76%) a complete agreement, in 24 cases (14%) a conditional agreement and in 18 patients (10%) no agreement were found. Assuming only stereotactic procedures the neuroradiological diagnosis was correct in 44 cases (66%) and incorrect in 10 cases (15%). In 13 patients (19%) one of the two differential diagnosis was correct. The specificity of the major tumors was calculated between 92% and 100%. The sensitivity for pituitary adenomas (n = 9) and neurinomas (n = 11) was 100%, the sensitivity for meningiomas (n = 32) was 94%. A sensitivity was calculated between 50% and 71% for astrocytomas (WHO I to WHO IV, n = 64) and metastases (n = 24). CONCLUSION: The accuracy was found to be higher than in the comparable retrospective studies.