Six endoscopic fenestrations of the 3rd ventricular floor have been performed in patients with stenosis (SAS) of the aqueduct of Sylvius in our institute during the last two years. The endoscopic intraventricular landmarks were the Monro's foramen followed by the mamillary bodies. The fenestration instrument was a monopolar coagulation wire, the dilatation instrument was a balloon catheter. The patients included two newborns and four adults. The two newborns developed a recurrent hydrocephalus after 2 months. The four adults remained well after the operation. The only complication was edema (SIADH syndrome) in one case for 24 hours. Flow sensitised phase MRI showed a mirroring in the prestenotic CSF pulsation curve preoperatively. This, in combination with an increased intraventricular pulsation, is a sign of reduced capacity of the subarachnoid space at the cerebral surface. The postoperative patency of the fenestration with diminished intraventricular pulsation can be demonstrated with ECG retrogated phase MRI. There was a slow and incomplete decrease of the preoperative enlarged ventricular size. This operative method is a low-risk, minimal invasive alternative method to shunt implantation in adults with SAS.

The experience of 7 operated patients with cavernous haemangiomas (CHa) and of 2 conservatively treated older patients is reported. There was no further postoperative neurological deficit, although 6 of the 7 patients had the CHa in an eloquent cerebral region. Two of the operated patients (22%) had several bleedings before surgery. In these cases seizures and visual field deficits remained. The relatively low rate of complications in our patient group was possible because an exact preoperative localisation helped to avoid a large traumatisation. This is possible with a stereotactic system or directly by CT guided skin marking. Intraoperative ultrasound was necessary in all cases because there was no landmark at the cerebral surface. The CHa was removed by a transsulcal microsurgical operation. A complete removal of the surrounding haemosiderin rim around the angioma seems necessary to avoid further seizures. The used technique has proved to be a simple and safe minimal invasive method.

Fifteen patients with solid and cystic occlusions of the cerebrospinal fluid (CSF) circulation pathways were examined with MRI using an ECG retrogated two-dimensional FISP sequence to determine whether there are certain defined pathological CSF flow patterns in these patients. All patients were clinically still compensated. The results of the measurements of CSF flow velocities at different locations in the CSF system were compared with the results from 8 healthy volunteers. In all patients with occlusive processes of the intraventricular CSF pathways (4 aqueduct stenoses, 1 membranous occlusion of the fourth ventricular outlets, 1 solid tumour at the foramen of Monro and 1 solid fourth ventricular tumour) we observed hyperdynamic CSF pulsation above the lesion. This was transmitted into the spinal canal. Close by a solid occlusion within the aqueduct the flow velocity curve over the RR cycle was typically shifted, resulting in a mirroring of the flow velocity curve compared with normals. In cystic lesions (n = 4) there was transmission of the pulsation wave through the lesion and therefore no mirroring of the flow velocity curve. This technique allows very good delineation of cystic structures in the peripineal region (n = 4), also due to the opposite direction of flow within the cyst compared with the surrounding CSF spaces, depending on the extent of communication.

A group of 318 patients suffering from severe head injury, 51 of whom died of their (head) injury, is reviewed. A small contusional cerebral haemorrhage had occurred in 50.6%, and 30% had had basal skull fractures. In the entire group, 2.8% developed posttraumatic hydrocephalus. Shunt implantation was necessary between the 11th and 123rd posttraumatic days (mean 52nd). All patients with inflammatory problems, e.g. meningitis or intracranial abscess, developed posttraumatic hydrocephalus. In 10 of the 12 patients with hygromas, an operative procedure with bur hole trepanation was necessary. Half of the patients with hygromas were polytraumatized but only 25.85% of the entire group. Two patients in the group and a further 49 patients with chronic subdural haematomas were operated on by a bur hole trepanation and appropriate drainage for several days. In 2 cases the development of the chronic subdural haematoma had been documented by CAI. All patients with frontobasal skull fractures were operated on by a transethmoidal approach only the 1st day. Only 2 of the laterobasal skull fractures needed an operation. One of these caused was by a pneumocephalus. In 0.9% of the patients an intracranial abscess was present; these all had a basal skull fracture, 1 after gunshot injury. Among the patients with basal skull fractures 3.1% developed carotid artery-cavernous sinus fistula. One of these three fistulas has been documented as yielding to spontaneous closure. After reangiography for balloon occlusion an aneurysm was found. Following primary operation, 2.5% of patients required revision operation because of renewed bleeding. The posttraumatic treatment of the complications is discussed.

36 patients with pituitary adenomas were examined via MRI to describe physiological changes and to visualise a residual tumour. Pre- and postoperative examinations included T1- and T2-weighted SE sequences. T1-weighted images were obtained in sagittal and coronal orientation pre- and post-Gd-DTPA application and T2-weighted images in coronal orientation. In 12 cases a residual tumour was found. Its signal intensity and contrast enhancement were similar to those of the primary tumour. Implanted material could be distinguished by localisation, decrease in volume and different signal intensity. The behaviour of contrast enhancement was helpful, since implanted material showed a rim enhancement. In our experience a sensitive imaging protocol in the follow-up of operated pituitary adenomas would be an early examination three months postoperatively followed by a control examination after one year. Information on the size and localisation of the primary tumour and the performed operative procedure is essential.

36 patients with pituitary adenomas were examined via MRI to describe physiological changes and to visualise a residual tumour. Pre- and postoperative examinations included T1- and T2-weighted SE sequences. T1-weighted images were obtained in sagittal and coronal orientation pre- and post-Gd-DTPA application and T2-weighted images in coronal orientation. In 12 cases a residual tumour was found. Its signal intensity and contrast enhancement were similar to those of the primary tumour. Implanted material could be distinguished by localisation, decrease in volume and different signal intensity. The behaviour of contrast enhancement was helpful, since implanted material showed a rim enhancement. In our experience a sensitive imaging protocol in the follow-up of operated pituitary adenomas would be an early examination three months postoperatively followed by a control examination after one year. Information on the size and localisation of the primary tumour and the performed operative procedure is essential.

16 patients with hydrocephalus communicans and 5 healthy volunteers were examined to demonstrate the pattern of the pulsatile CSF flow. After implantation of a CSF shunt system the same patients were examined again to show the influence of the shunt on the CSF pulsations. We used a flow-sensitised, cardiac-gated 2D FLASH sequence and analysed the phase and magnitude images. It could be shown that most patients (n=12) had a hyperdynamic pulsatile flow preoperatively. After shunt implantation the pulsatile CSF motion and the clinical symptoms were improved in 8 of these patients. MRI of pulsatile CSF flow movement seems to be a helpful noninvasive tool to estimate the prognosis of a shunt implantation in patients with hydrocephalus communicans.

16 patients with hydrocephalus communicans and 5 healthy volunteers were examined to demonstrate the pattern of the pulsatile CSF flow. After implantation of a CSF shunt system the same patients were examined again to show the influence of the shunt on the CSF pulsations. We used a flow-sensitised, cardiac-gated 2D FLASH sequence and analysed the phase and magnitude images. It could be shown that most patients (n = 12) had a hyperdynamic pulsatile flow preoperatively. After shunt implantation the pulsatile CSF motion and the clinical symptoms were improved in 8 of these patients. MRI of pulsatile CSF flow movement seems to be a helpful noninvasive tool to estimate the prognosis of a shunt implantation in patients with hydrocephalus communicans.