COMPUTER-AIDED LESION DETECTION WITH STATISTICAL MODEL-BASED FEATURESIN PET IMAGES

Positron emission tomography (PET) with the glucose analog [F-18] fluo rodeoxyglucose is proving to be useful in cancer diagnosis and treatme nt. However, as in all nuclear medicine imaging technologies, lesion d etection with PET is often hindered by limited spatial resolution and low signal-to-noise ratios. Under such conditions, conventional diagno sis by visual inspection usually becomes difficult and potentially ina ccurate. In this paper, we propose use of computer-aided lesion detect ion methods for PET imaging by applying a maximum likelihood ratio tes t and a composite hypothesis test, assuming that the mean positron emi ssion rate is deterministic or random, respectively. In our approach, different statistical models characterizing the mean positron emission rate, the raw sinogram data and the filtered backprojection (FBP) rec onstructed image are used to derive the test criteria. Three methods t o estimate the unknown parameters of the test functions from observati ons are presented. The performance of one of the proposed methods is e valuated and compared with both simulated and experimental phantom dat a. In the preliminary trials, the methods detect correctly (with a hig h probability > 0.9) lesions of diameter greater than or equal to 15mm with lesion-to-background contrast 1.1:1. Under the same conditions, the test lesion could not be detected by visual inspection alone in th e images reconstructed by either the FBP or the maximum likelihood ite rative algorithms. The methods may also be used for the objective asse ssment of the quality of images reconstructed from different algorithm s.