210 SOCIETE D'E.E.G. ET DE NEUROPHYSIOLOGIE CLINIQUE DE LANGUE FRAN~A|SE Regional cerebral blood flow in cerebral infarction and in transient ischemic attacks. Olaf B. PAULSON. Department of Neurology and Clinical Physiology, Bispebierx~ Hospital, DK-2400 Copenhagen, Denmark. The ~3aXenon intra-arterial injection method allows the simultaneous determination of blood flow in multiple areas of the brain. This method has during recent years been used in several series of vascular diseases of the brain, and focal flow abnormalities, including focal vasomotor paralysis, have been reported. In addition, disturbed vasomotor function may also be observed outside the focal area and even in the hemisphere contralateral to the focal lesion. These studies of cerebrovas- cular diseases have recently been reviewed in detail (11 ). The present paper summarizes the ~:~3 Xenon inna-arterial injection method and the clinical results in vascular diseases of the brain. THB 133XENON INTRA-AR1ERIAI, INJECTION METIIOD. A small polyethylene catheter is placed in the internal carotid artery by means of'the Seldinger technique. Two to three mCi of~3Z~Xenon dissolved in one to three ml of isotonic saline is injected rapidly (one to two seconds) through the catheter and the clearance of the isotope is followed by multiple small collimated scintillation detectors placed externally over the ipsilateral hemisphere (NaI(TI) crystals, 12 mm in diameter and 5 or 10 mm thick) (fig. I), During each flow measurement a blood sample is drawn for the determination of the PaCO~ and the intracarotid blood pressure is measured using an electric manometer. After the flow study an angiographic examination is performed through the indwelling catheter. The counting fields are represented by truncated cones through the hemisphere ([2). The counting fields from neighboring channels overlap, especially in the medial parts of tile hemisphere (fig. 1). Therefore, in case of focal cerebral diseases, even in tile cotinting field of the detectors overlying the focal lesion, there will be an overlapping of diseased and non diseased cerebral tissue. Thus, the flow in the focal lesion itself carmot be accurately determined. Compton scatter will further reduce the resolution somewhat. The average flow measured from a region, can be calculated from the isotope clearance curve recorded over a 10 or 15 minutes period (6). Usually a l0 minutes clearance period is clsed and the calculated flow vaIues represent an overestimation of about 15 % of the real average flow (9). The calculation utilizes the same concepts as those underlying the classical Kety-Schmidt inert gas technique (7). It is also possible to calculate the flow of the grey and white matter separately using a two compartmental analysis (4, 6). In cases with focal cerebral diseases, however, the two compartmental model, most often brakes down (5, 10). The regional cerebral blood flow can also be calculated from the initial part of the ~:~Xenon clearance curve (the first two minutes) (5, 9, 12). This calculation gives flow values which are approximations of the flow in the grey matter, the fIow calculated from the initial part of tlle clearance curve beeing about 25 o//o lower than the flow in the grey matter (9). The flow value calculated from the initial part of the clearance curve has been called rCBF inilial and is calculated as : CBF initial 2 × D initial ml/100 g/rain. 2 being the product of the conversion factor 2.30 (ln 10) from natural to-10 logarithm, and of the tissue-to- blood partition coefficient for ~33Xenon taken to be 0.87.D initial is the initial slope of the logarithmically recorded clearance curve in per cent of decade per minute. The flow values may be corrected for remaining activity from previous t'~aXenon injection (9). It has been demonstrated that rCBF initial is highly reproducible (9) and most suitable for the determi- nation of local flow abnormalities and disturbed vasomotor function (5, 8, 10, 12, 13, 18). The logarithmi- cally recorded clearance curve, from which the CBF inilial is calculated will normally decrease in an essen Tir~s D part : O. B. PAULSON (A I'adresse ci-dessus).