Speaker: Dan J. Kadrmas

Title: Evaluation of Multiclass Model Observers in PET LROC Studies

Topic: I will be discussing model observers for evaluating lesion detectability of PET images, and will discuss them in the context of the following paper:

Reference: Evaluation of Multiclass Model Observers in PET LROC Studies H.C. Gifford, P.E. Kinahan, C. Lartizien, and M.A. King IEEE Trans. Nucl. Sci 51(1):116-123, 2007.

Abstract: A localization ROC (LROC) study was conducted to evaluate nonprewhitening matched-filter (NPW) and chan- nelized NPW (CNPW) versions of a multiclass model observer as predictors of human tumor-detection performance with PET images. Target localization is explicitly performed by these model observers. Tumors were placed in the liver, lungs, and background soft tissue of a mathematical phantom, and the data simulation modeled a full-3D acquisition mode. Reconstructions were per- formed with the FORE AWOSEM algorithm. The LROC study measured observer performance with 2D images consisting of either coronal, sagittal, or transverse views of the same set of cases. Versions of the CNPW observer based on two previously published difference-of-Gaussian channel models demonstrated good quantitative agreement with human observers. One interpre- tation of these results treats the CNPW observer as a channelized Hotelling observer with implicit internal noise.