EVALUATION OF A CROSS-VALIDATION STOPPING RULE IN MLE SPECT RECONSTRUCTION

One of the problems in the routine use of the maximum-likelihood estim ator method-expectation maximization (MLE-EM) algorithm is to decide w hen the iterative process should be stopped. We studied a cross-valida tion stopping rule to assess its usefulness in SPECT. We tested this s topping rule criterion in the MLE-EM algorithm without acceleration as well as in two accelerating algorithms, the successive substitutions algorithm (SSA) and the additive algorithm (AA). Different values of a n acceleration factor were tested in SSA and AA. Our results from nume rical and physical phantoms show that the stopping rule based on the c ross-validation ratio (CVR) takes into account the similarity of the r econstructed image to the ideal image, noise and the contrast of the i mage. CVR yields reconstructed images with balanced values of the figu res of merit (FOM) employed to assess the image quality. The CVR crite rion can be used in the original MLE-EM algorithm as well as in SSA an d AA. The reconstructed images obtained with SSA and AA showed FOM val ues that were very similar. These results were justified by considerin g AA to be an approximate form of SSA. The range of validity for the a cceleration factor in SSA and AA was found to be [1,2]. In this range, an inverse function connects the acceleration factor to the number of iterations needed to attain prefixed values of FOMs.