Diagnostic utility of artificial intelligence for left ventricular scar identification using cardiac magnetic resonance imaging: A systematic review

Nikesh Jathanna; Anna Podlasek; Albert Sokol; Dorothee Auer; Xin Chen; Shahnaz Jamil-Copley

doi:10.1016/j.cvdhj.2021.11.005

Diagnostic utility of artificial intelligence for left ventricular scar identification using cardiac magnetic resonance imaging: A systematic review

Nikesh Jathanna (Lead / Corresponding author), Anna Podlasek, Albert Sokol, Dorothee Auer, Xin Chen, Shahnaz Jamil-Copley

Research output: Contribution to journal › Article › peer-review

10 Citations (Scopus)

6 Downloads (Pure)

Abstract

Background: Accurate, rapid quantification of ventricular scar using cardiac magnetic resonance imaging (CMR) carries importance in arrhythmia management and patient prognosis. Artificial intelligence (AI) has been applied to other radiological challenges with success.

Objective: We aimed to assess AI methodologies used for left ventricular scar identification in CMR, imaging sequences used for training, and its diagnostic evaluation.

Methods: Following PRISMA recommendations, a systematic search of PubMed, Embase, Web of Science, CINAHL, OpenDissertations, arXiv, and IEEE Xplore was undertaken to June 2021 for full-text publications assessing left ventricular scar identification algorithms. No pre-registration was undertaken. Random-effect meta-analysis was performed to assess Dice Coefficient (DSC) overlap of learning vs predefined thresholding methods.

Results: Thirty-five articles were included for final review. Supervised and unsupervised learning models had similar DSC compared to predefined threshold models (0.616 vs 0.633, P = .14) but had higher sensitivity, specificity, and accuracy. Meta-analysis of 4 studies revealed standardized mean difference of 1.11; 95% confidence interval -0.16 to 2.38, P = .09, I² = 98% favoring learning methods.

Conclusion: Feasibility of applying AI to the task of scar detection in CMR has been demonstrated, but model evaluation remains heterogenous. Progression toward clinical application requires detailed, transparent, standardized model comparison and increased model generalizability.

Original language	English
Pages (from-to)	S21-S29
Number of pages	9
Journal	Cardiovascular Digital Health Journal
Volume	2
Issue number	6
Early online date	24 Nov 2021
DOIs	https://doi.org/10.1016/j.cvdhj.2021.11.005
Publication status	Published - Dec 2021

Keywords

Artificial intelligence
Cardiac scar
Deep learning
Imaging – cardiac magnetic resonance imaging (MRI)
Machine learning
Neural networks

ASJC Scopus subject areas

Biomedical Engineering
Critical Care and Intensive Care Medicine
Cardiology and Cardiovascular Medicine

Access to Document

10.1016/j.cvdhj.2021.11.005Licence: CC BY-NC-ND

Final published versionFinal published version, 170 KBLicence: CC BY-NC-ND

Cite this

@article{cfa017797d3b4c068354b78b6f513451,

title = "Diagnostic utility of artificial intelligence for left ventricular scar identification using cardiac magnetic resonance imaging: A systematic review",

abstract = "Background: Accurate, rapid quantification of ventricular scar using cardiac magnetic resonance imaging (CMR) carries importance in arrhythmia management and patient prognosis. Artificial intelligence (AI) has been applied to other radiological challenges with success. Objective: We aimed to assess AI methodologies used for left ventricular scar identification in CMR, imaging sequences used for training, and its diagnostic evaluation. Methods: Following PRISMA recommendations, a systematic search of PubMed, Embase, Web of Science, CINAHL, OpenDissertations, arXiv, and IEEE Xplore was undertaken to June 2021 for full-text publications assessing left ventricular scar identification algorithms. No pre-registration was undertaken. Random-effect meta-analysis was performed to assess Dice Coefficient (DSC) overlap of learning vs predefined thresholding methods. Results: Thirty-five articles were included for final review. Supervised and unsupervised learning models had similar DSC compared to predefined threshold models (0.616 vs 0.633, P = .14) but had higher sensitivity, specificity, and accuracy. Meta-analysis of 4 studies revealed standardized mean difference of 1.11; 95% confidence interval -0.16 to 2.38, P = .09, I2 = 98% favoring learning methods. Conclusion: Feasibility of applying AI to the task of scar detection in CMR has been demonstrated, but model evaluation remains heterogenous. Progression toward clinical application requires detailed, transparent, standardized model comparison and increased model generalizability.",

keywords = "Artificial intelligence, Cardiac scar, Deep learning, Imaging – cardiac magnetic resonance imaging (MRI), Machine learning, Neural networks",

author = "Nikesh Jathanna and Anna Podlasek and Albert Sokol and Dorothee Auer and Xin Chen and Shahnaz Jamil-Copley",

note = "Publisher Copyright: {\textcopyright} 2021 Heart Rhythm Society",

year = "2021",

month = dec,

doi = "10.1016/j.cvdhj.2021.11.005",

language = "English",

volume = "2",

pages = "S21--S29",

journal = "Cardiovascular Digital Health Journal",

issn = "2666-6936",

publisher = "Elsevier",

number = "6",

}

Diagnostic utility of artificial intelligence for left ventricular scar identification using cardiac magnetic resonance imaging: A systematic review. / Jathanna, Nikesh (Lead / Corresponding author); Podlasek, Anna; Sokol, Albert et al.
In: Cardiovascular Digital Health Journal, Vol. 2, No. 6, 12.2021, p. S21-S29.

Research output: Contribution to journal › Article › peer-review

TY - JOUR

T1 - Diagnostic utility of artificial intelligence for left ventricular scar identification using cardiac magnetic resonance imaging

T2 - A systematic review

AU - Jathanna, Nikesh

AU - Podlasek, Anna

AU - Sokol, Albert

AU - Auer, Dorothee

AU - Chen, Xin

AU - Jamil-Copley, Shahnaz

PY - 2021/12

Y1 - 2021/12

N2 - Background: Accurate, rapid quantification of ventricular scar using cardiac magnetic resonance imaging (CMR) carries importance in arrhythmia management and patient prognosis. Artificial intelligence (AI) has been applied to other radiological challenges with success. Objective: We aimed to assess AI methodologies used for left ventricular scar identification in CMR, imaging sequences used for training, and its diagnostic evaluation. Methods: Following PRISMA recommendations, a systematic search of PubMed, Embase, Web of Science, CINAHL, OpenDissertations, arXiv, and IEEE Xplore was undertaken to June 2021 for full-text publications assessing left ventricular scar identification algorithms. No pre-registration was undertaken. Random-effect meta-analysis was performed to assess Dice Coefficient (DSC) overlap of learning vs predefined thresholding methods. Results: Thirty-five articles were included for final review. Supervised and unsupervised learning models had similar DSC compared to predefined threshold models (0.616 vs 0.633, P = .14) but had higher sensitivity, specificity, and accuracy. Meta-analysis of 4 studies revealed standardized mean difference of 1.11; 95% confidence interval -0.16 to 2.38, P = .09, I2 = 98% favoring learning methods. Conclusion: Feasibility of applying AI to the task of scar detection in CMR has been demonstrated, but model evaluation remains heterogenous. Progression toward clinical application requires detailed, transparent, standardized model comparison and increased model generalizability.

AB - Background: Accurate, rapid quantification of ventricular scar using cardiac magnetic resonance imaging (CMR) carries importance in arrhythmia management and patient prognosis. Artificial intelligence (AI) has been applied to other radiological challenges with success. Objective: We aimed to assess AI methodologies used for left ventricular scar identification in CMR, imaging sequences used for training, and its diagnostic evaluation. Methods: Following PRISMA recommendations, a systematic search of PubMed, Embase, Web of Science, CINAHL, OpenDissertations, arXiv, and IEEE Xplore was undertaken to June 2021 for full-text publications assessing left ventricular scar identification algorithms. No pre-registration was undertaken. Random-effect meta-analysis was performed to assess Dice Coefficient (DSC) overlap of learning vs predefined thresholding methods. Results: Thirty-five articles were included for final review. Supervised and unsupervised learning models had similar DSC compared to predefined threshold models (0.616 vs 0.633, P = .14) but had higher sensitivity, specificity, and accuracy. Meta-analysis of 4 studies revealed standardized mean difference of 1.11; 95% confidence interval -0.16 to 2.38, P = .09, I2 = 98% favoring learning methods. Conclusion: Feasibility of applying AI to the task of scar detection in CMR has been demonstrated, but model evaluation remains heterogenous. Progression toward clinical application requires detailed, transparent, standardized model comparison and increased model generalizability.

KW - Artificial intelligence

KW - Cardiac scar

KW - Deep learning

KW - Imaging – cardiac magnetic resonance imaging (MRI)

KW - Machine learning

KW - Neural networks

UR - http://www.scopus.com/inward/record.url?scp=85139185914&partnerID=8YFLogxK

U2 - 10.1016/j.cvdhj.2021.11.005

DO - 10.1016/j.cvdhj.2021.11.005

M3 - Article

AN - SCOPUS:85139185914

SN - 2666-6936

VL - 2

SP - S21-S29

JO - Cardiovascular Digital Health Journal

JF - Cardiovascular Digital Health Journal

IS - 6

ER -

Diagnostic utility of artificial intelligence for left ventricular scar identification using cardiac magnetic resonance imaging: A systematic review

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this