Events Calendar

Penalized Maximum-likelihood PET Image Reconstruction for Lesion Detection

Speaker: Li Yang, University of California-Davis

Talk Title: Penalized Maximum-likelihood PET Image Reconstruction for Lesion Detection

Abstract: Detecting cancerous lesions is a major clinical application in emission tomography. Statistical reconstruction methods based on the penalized maximum-likelihood (PML) principle have been developed to improve image quality. A number of metrics have been used to evaluate the quality of the reconstructed PET images, such as spatial resolution, noise variance, contrast-to-noise ratio, etc. Work has been done to optimize PML reconstruction to achieve uniform resolution and to maximize the contrast-to-noise ratio. However, these technical metrics do not necessarily reflect the performance of a clinical task. Here we focus on lesion detection and use a task-specific metric to evaluate the image quality. A multiview channelized Hotelling observer (mvCHO) is used to assess the lesion detectability in 3D images to mimic the condition where a human observer examines three orthogonal views of a 3D image for lesion detection. We derive simplified theoretical expressions that allow fast prediction of the detectability of a 3D lesion. We apply the theoretical results to guide the design of a shift-variant quadratic penalty function in PML reconstruction to maximize detectability of lesions at unknown locations in fully 3D PET. The proposed method is evaluated using computer-based Monte Carlo simulations as well as real patient data with a superimposed lesion.

Furthermore, we extend our theoretical analysis of static PET reconstruction to dynamic PET. We study both the conventional indirect reconstruction and direct reconstruction for Patlak parametric image estimation. In indirect reconstruction, Patlak parametric images are generated by reconstructing a sequence of dynamic PET images first and then performing Patlak analysis on the time activity curves (TACs) pixel-by-pixel. In direct reconstruction, Patlak parametric images are estimated directly from raw sinogram data by incorporating the Patlak model into the image reconstruction procedure. The PML reconstruction is used in both the indirect and direct reconstruction methods. Simplified expressions for evaluating the lesion detectability on Patlak parametric images have been derived and applied to the selection of the regularization parameter value to maximize the lesion detectability. Good agreements between the theoretical predictions and the Monte Carlo results are observed. The theoretical formula also shows the benefit of the direct method in dynamic PET reconstruction for lesion detection.


Biography: Li Yang received his B.S. degree in precision instrumentation from Tianjin University (China) in 2009. Currently, he is pursuing his Ph.D. degree in biomedical engineering at University of California-Davis under the supervision of Prof. Jinyi Qi. His research interests are image quality evaluation and statistical image reconstruction for emission tomography


Host: Prof. Richard Leahy

Location: Hughes Aircraft Electrical Engineering Center (EEB) - 25

Audiences: Everyone Is Invited

Contact: Talyia Veal