Measurement of the arterial input function using a non-invasive wrist scanner

556 Objectives: The purpose of this study is to validate a wrist scanner which will eliminate the need for invasive sampling of arterial blood to determine an input function (a measure of radioactivity in blood as a function of time). Use of this device will minimize the discomfort to the patient and risk to medical personnel. Methods: The prototype is a compact 4 detector block system. Each block consists of a 4×8 array of 32 2×2×15mm3 LSO crystals and Avalanche Photodiodes. Efficiency, spatial resolution and temporal response of the system were measured using an anatomically correct wrist phantom. High resolution planar images were generated to determine if the artery can be visually separated from the vein. MicroPET images of a human wrist were obtained to determine if activity in the surrounding tissue would cause an unacceptable loss of signal for the artery. The scanner was then used in a PET study, where a human subject was injected with 6.1mCi of 11C Raclopride. Focal plane reconstruction method was used to generate planar images and a time activity curve was derived. Results: Data from the phantom studies demonstrated that when a solution containing 2μCi/cc was passed through the tube representing the artery, an efficiency of 4cps/nCi/cc was obtained. The MicroPET images show that the arterial signal is significantly higher than the contribution from the surrounding tissue. This background was estimated to contribute about 12% to the total observed arterial activity determined by drawing ROIs on the wrist image. The input function from these ROIs matched well with the invasive samples. The human scan demonstrated the feasibility of using the scanner in a clinical setting. Human scans will continue to optimize the placement of the detectors, and to determining both the amount and best placement of shielding around the wrist scanner to reduce background counts. The image derived input function from the wrist scanner will be compared to the invasive samples. Conclusions: The wrist scanner will enable noninvasive measurement of the input function, thereby eliminating invasive blood sampling. The next prototype will have 8 detectors. Monte Carlo based simulations have shown that this will increase the sensitivity by a factor of 3.4. Research Support (if any): DOE DE-AC02-98CH10886