Uzaysal düzenlileştirme ile kestirilen PET parametrelerindeki yanlilik ve deǧişinti analizi

Translated title of the contribution: Bias and variance analysis of PET parameters estimated with spatial regularization

Mustafa E. Kamasak*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this work, an analytical framework is developed for bias and variance in the kinetic parameter estimations with spatial regularization in dynamic positron emission tomography. Time consuming Monte Carlo simulations can be used for this purpose. In this work, a faster analytical framework is developed for bias and variance analysis in kinetic parameter estimations with spatial regularization. In addition validation experiments are performed on simulation data. It is observed that the bias and variance values obtained from Monte Carlo simulations and analytical calculations are consistent. These results indicate that the bias and variance in the kinetic parameter estimations with spatial regularization can be computed using the analytical framework that is derived in this work.

Translated title of the contributionBias and variance analysis of PET parameters estimated with spatial regularization
Original languageTurkish
Title of host publication2013 21st Signal Processing and Communications Applications Conference, SIU 2013
DOIs
Publication statusPublished - 2013
Event2013 21st Signal Processing and Communications Applications Conference, SIU 2013 - Haspolat, Turkey
Duration: 24 Apr 201326 Apr 2013

Publication series

Name2013 21st Signal Processing and Communications Applications Conference, SIU 2013

Conference

Conference2013 21st Signal Processing and Communications Applications Conference, SIU 2013
Country/TerritoryTurkey
CityHaspolat
Period24/04/1326/04/13

Fingerprint

Dive into the research topics of 'Bias and variance analysis of PET parameters estimated with spatial regularization'. Together they form a unique fingerprint.

Cite this