Classification of Myocardial
Computer Aided Diagnosis
Computer Application-Detection/Diagnosis
Heart
PET
Journal
Radiology. Artificial intelligence
ISSN: 2638-6100
Titre abrégé: Radiol Artif Intell
Pays: United States
ID NLM: 101746556
Informations de publication
Date de publication:
Jul 2021
Jul 2021
Historique:
received:
30
06
2020
revised:
17
02
2021
accepted:
11
03
2021
entrez:
5
8
2021
pubmed:
6
8
2021
medline:
6
8
2021
Statut:
epublish
Résumé
To perform automated myocardial segmentation and uptake classification from whole-body fluorine 18 fluorodeoxyglucose (FDG) PET. In this retrospective study, consecutive patients who underwent FDG PET imaging for oncologic indications were included (July-August 2018). The left ventricle (LV) on whole-body FDG PET images was manually segmented and classified as showing no myocardial uptake, diffuse uptake, or partial uptake. A total of 609 patients (mean age, 64 years ± 14 [standard deviation]; 309 women) were included and split between training (60%, 365 patients), validation (20%, 122 patients), and testing (20%, 122 patients) datasets. Two sequential neural networks were developed to automatically segment the LV and classify the myocardial uptake pattern using segmentation and classification training data provided by human experts. Linear regression was performed to correlate findings from human experts and deep learning. Classification performance was evaluated using receiver operating characteristic (ROC) analysis. There was moderate agreement of uptake pattern between experts and deep learning (as a fraction of correctly categorized images) with 78% (36 of 46) for no uptake, 71% (34 of 48) for diffuse uptake, and 71% (20 of 28) for partial uptake. There was no bias in LV volume for partial or diffuse uptake categories Deep learning was able to segment and classify myocardial uptake patterns on FDG PET images.
Identifiants
pubmed: 34350405
doi: 10.1148/ryai.2021200148
pmc: PMC8328107
doi:
Types de publication
Journal Article
Langues
eng
Pagination
e200148Subventions
Organisme : NIBIB NIH HHS
ID : T32 EB009384
Pays : United States
Informations de copyright
2021 by the Radiological Society of North America, Inc.
Déclaration de conflit d'intérêts
Disclosures of Conflicts of Interest: N.J. Activities related to the present article: institution receives grant (1R01HL137984-01A1). Activities not related to the present article: disclosed no relevant relationships. Other relationships: disclosed no relevant relationships. M.T.M. Activities related to the present article: author received grant from Sarnoff Cardiovascular Research Foundation. Activities not related to the present article: disclosed no relevant relationships. Other relationships: disclosed no relevant relationships. C.J. disclosed no relevant relationships. B.F. Activities related to the present article: given minimum wage through grant money. Activities not related to the present article: disclosed no relevant relationships. Other relationships: disclosed no relevant relationships. B.F.M. Activities related to the present article: paid consulting fee or honorarium by University of Pennsylvania. Activities not related to the present article: disclosed no relevant relationships. Other relationships: disclosed no relevant relationships. E.H. disclosed no relevant relationships. S.K.I. disclosed no relevant relationships. H.L. disclosed no relevant relationships. Y.H. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author was consultant for GE Healthcare; institution has grants from Gilead and Pfizer. Other relationships: disclosed no relevant relationships. F.K. disclosed no relevant relationships. P.E.B. disclosed no relevant relationships. W.R.W. Activities related to the present article: institution received grant from National Institutes of Health/National Heart, Lung, and Blood Institute (1R01HL137984-01A1). Activities not related to the present article: disclosed no relevant relationships. Other relationships: disclosed no relevant relationships.
Références
Ann Nucl Med. 2013 Jul;27(6):572-80
pubmed: 23546808
Lancet. 1963 Apr 13;1(7285):785-9
pubmed: 13990765
Nature. 2017 Feb 2;542(7639):115-118
pubmed: 28117445
Sci Rep. 2020 Mar 24;10(1):5331
pubmed: 32210327
Phys Med Biol. 2018 Jul 16;63(14):145011
pubmed: 29923839
J Nucl Cardiol. 2016 Apr;23(2):244-52
pubmed: 26243179
Hum Brain Mapp. 2018 Sep;39(9):3769-3778
pubmed: 29752765
J Nucl Cardiol. 2009 Sep-Oct;16(5):801-10
pubmed: 19548047
J Nucl Med. 2018 Oct;59(10):1624-1629
pubmed: 29449446
IEEE Trans Biomed Eng. 2017 Jan;64(1):134-144
pubmed: 27046887
Annu Rev Physiol. 1974;36:413-59
pubmed: 19400669
Phys Med. 2015 Dec;31(8):969-980
pubmed: 26321409
Ann Nucl Med. 1998 Apr;12(2):95-103
pubmed: 9637280
AMIA Annu Symp Proc. 2018 Dec 05;2018:1571-1580
pubmed: 30815203
J Med Invest. 2014;61(1-2):53-8
pubmed: 24705749
JAMA. 2016 Dec 13;316(22):2402-2410
pubmed: 27898976
Nucl Med Biol. 2007 Oct;34(7):737-42
pubmed: 17921026
J Nucl Med. 2004 Dec;45(12):1989-98
pubmed: 15585472
Clin Nucl Med. 2016 Jul;41(7):e327-39
pubmed: 26646995
Eur J Nucl Med Mol Imaging. 2019 Jun;46(6):1240-1247
pubmed: 30327855
Nature. 2018 Oct;562(7726):203-209
pubmed: 30305743
JACC Cardiovasc Imaging. 2020 Apr;13(4):1096-1097
pubmed: 31954654
J Nucl Med. 2007 Oct;48(10):1662-9
pubmed: 17873124
Phys Med. 2019 Feb;58:32-39
pubmed: 30824147
Eur J Nucl Med. 1992;19(10):874-81
pubmed: 1451704