Calibration of detector sensitivities in a cylindrical PET camera

A method for calibrating the detectors in a cylindrical positron camera that uses cross-plane data is presented, and its performance is described. The method is based on the assumption that the sensitivity of a pair of detectors is the product of the individual sensitivity of the two detectors and a geometric factor. The calibration source is an annulus or rotating line of activity close to the detectors. The method uses measurements of the sum of all the coincidences from a detector, solves for geometric factors and sensitivities simultaneously, and is iterative. A key part of the method is a technique for forcing the iteration to converge quickly, usually within two or three iterations. The precision of the estimate of a detector´s sensitivity approaches the theoretical lower bound determined by the sum of all the counts in the fan of coincidences from that detector