Prediction of Standard-Dose PET Image by Low-Dose PET and MRI Images

Jiayin Kang, Yaozong Gao, Yao Wu, Guangkai Ma, Feng Shi, Weili Lin, Dinggang Shen

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Positron emission tomography (PET) is a nuclear medical imaging technology that produces 3D images of tissue metabolic activity in human body. PET has been used in various clinical applications, such as diagnosis of tumors and diffuse brain disorders. High quality PET image plays an essential role in diagnosing diseases/disorders and assessing the response to therapy. In practice, in order to obtain the high quality PET images, standard-dose radionuclide (tracer) needs to be used and injected into the living body. As a result, it will inevitably increase the risk of radiation. In this paper, we propose a regression forest (RF) based framework for predicting standard-dose PET images using low-dose PET and corresponding magnetic resonance imaging (MRI) images instead of injecting the standard-dose radionuclide into the body. The proposed approach has been evaluated on a dataset consisting of 7 subjects using leave-one-out cross-validation. Moreover, we compare the prediction performance between sparse representation (SR) based method and our proposed method. Both qualitative and quantitative results illustrate the practicability of our proposed method.

Original languageEnglish
Pages (from-to)280-288
Number of pages9
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8679
Publication statusPublished - 2014 Jan 1
Externally publishedYes

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Fingerprint Dive into the research topics of 'Prediction of Standard-Dose PET Image by Low-Dose PET and MRI Images'. Together they form a unique fingerprint.

  • Cite this