Computerized Tomography (CT) has acquired a preeminent position in the neuroradiological diagnosis and already played an important role in the neurosurgical practice. The interpretation of CT scans is commonly based on the CT images of cathode ray tube. However, CT essentially provides us the informations in digital form and stores them on the disc and/or the magnetic tape.
The authors have attempted to analyse the data stored on the magnetic tape of EMI scanner (Model 1000) with another computer system (FACOM 230/45S) under three different programs, that is, histogram, profile and isometric display. The objects of analysis were data acquired from the contrast studies in histologically verified cases of intracranial tumors involving 61 cases of cerebral neoplasm and 9 cases of the other mass lesion. The results were as follows :
1) When change of pre and post-contrast attenuation values were within 1.0±0.3 of EMI-number, the difference can be regarded as non-significant.
2) Most of preferential tumors in the cerebral hemisphere, the cerebello-pontine angle and the parasellar region, have possessed characteristic attenuation values and revealed certain enhancement effects.
3) The histographic and the isometric displays have expressed characteristic patterns of the tumor types particularly in the cases of malignant astrocytoma, meningioma, pituitary adenoma, and hemangioblastoma.
4) To evaluate homogeneity of the tumor tissue, the authors have proposed an index for the convenience on the analysis of histogram. It has been able to discriminate the solid tumors from the cystic tumors and/or the necrotizing tissues by analyzing these indices.
5) From the analysis of the profile display, different patterns of ring-like enhancement were recognized between the cases of malignant astrocytoma and metastatic carcinoma, and the cases of cerebral abscess and resolving hematoma. So it was supposed that different mechanism should exist in the mechanism of ringlike image formation in different diseases.
From the above-described results, computer-assisted analysis of the EMI-number, in combination with the CRT images, could be greatfully contributory to differential diagnosis of the cerebral tumors in vivo. Moreover, it will be expected that these analytical methods could participate in further developments for the preoperative diagnosis of brain tumor in such a form of automatic diagnosis.
View full abstract