III. The Accuracy of Radiology in Demonstrating Ruptured Intracranial Aneurysms

210 cases of subarachnoid haemorrhage are reviewed. The cases were selected in that all suffered from a ruptured aneurysm, all were submitted to cerebral angiography and the diagnosis was confirmed at autopsy in every instance. The angiographic accuracy was 89 per cent. The causes of error in diagnosis in the remaining 11 per cent are discussed. The authors divided the errors into five groups: 1. Spasm preventing filling of the aneurysm by contrast substance. 2. Observer's error. 3. Inadequate examination; the most important short-coming being a paucity of projections taken at different angles. 4. Normal arteriograms even in retrospect after the most careful scrutiny. In every case (2 per cent) the aneurysm was small. 5. Intact aneurysm observed, second aneurysm (ruptured) not observed. In no instance did thrombus prevent the angiographic demonstration of an aneurysm and it is felt that thrombosis has been blamed too readily in the past for discrepancies in radiological and autopsy findings. The value of projections such as the per-orbital view in addition to routine antero-posterior, lateral and oblique views, is stressed. The authors believe that the potential accuracy of radiological diagnosis is approximately 96 per cent.