The Performance of Machine Learning Approaches for Attenuation Correction of PET in Neuroimaging: A Meta-Analysis - 03/02/23

Doi : 10.1016/j.neurad.2023.01.157

Raymond Confidence ^1,², Michael T. Jurkiewicz ^1,^2,³, Akin Orunmuyi ⁴, Linshan Liu ², Michael Oluwaseun Dada ⁵, Claes N. Ladefoged ⁶, Jarmo Teuho ^7,⁸, Udunna C Anazodo ^1,^2,^9,^⁎
¹ Department of Medical Biophysics, Western University, London, ON, Canada.
² Lawson Health Research Institute, London, ON, Canada
³ Department of Medical Imaging, Western University, London, ON, Canada.
⁴ Kenyatta University Teaching, Research and Referral Hospital, Nairobi, Kenya.
⁵ Department of Physics, Federal University of Technology, Minna, Nigeria.
⁶ Department of Clinical Physiology, Nuclear Medicine, and PET, Rigshospitalet Copenhagen, Denmark.
⁷ Turku PET Centre, Turku University, Turku, Finland.
⁸ Turku University Hospital, Turku, Finland.
⁹ Montreal Neurological Institute, Montreal, Quebec, Canada

^⁎Corresponding Author: 3801 Rue University, Montreal, QC H3A 2B43801 Rue UniversityMontrealQCH3A 2B4

Sous presse. Manuscrit accepté. Disponible en ligne depuis le Friday 03 February 2023
Cet article a été publié dans un numéro de la revue, cliquez ici pour y accéder

Abstract

Purpose

This systematic review provides a consensus on the clinical feasibility of machine learning (ML) methods for brain PET attenuation correction (AC). Performance of ML-AC were compared to clinical standards.

Methods

Two hundred and eighty studies were identified through electronic searches of brain PET studies published between January 1, 2008, and August 1, 2022. Reported outcomes for image quality, tissue classification performance, regional and global bias were extracted to evaluate ML-AC performance. Methodological quality of included studies and the quality of evidence of analysed outcomes were assessed using QUADAS-2 and GRADE, respectively.

Results

A total of 19 studies (2371 participants) met the inclusion criteria. Overall, the global bias of ML methods was 0.76 ± 1.2%. For image quality, the relative mean square error (RMSE) was 0.20 ± 0.4 while for tissues classification, the Dice similarity coefficient (DSC) for bone/soft tissue/air were 0.82 ± 0.1 / 0.95 ± 0.03 / 0.85 ± 0.14.

Conclusions

In general, ML-AC performance is within acceptable limits for clinical PET imaging. The sparse information on ML-AC robustness and its limited qualitative clinical evaluation may hinder clinical implementation in neuroimaging, especially for PET/MRI or emerging brain PET systems where standard AC approaches are not readily available.

Le texte complet de cet article est disponible en PDF.