Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation: A Multi-Site Study

Golnoush Asaeikheybari; Majd El-Harasis; Amit Gupta; M. Benjamin Shoemaker; John Barnard; Joshua Hunter; Rod S. Passman; Han Sun; Hyun Su Kim; Taylor Schilling; William Telfer; Britta Eldridge; Po Hao Chen; Abhishek Midya; Bibin Varghese; Samuel J. Harwood; Alison Jin; Sojin Y. Wass; Aleksandar Izda; Kevin Park; Abel Abraham; David R. Van Wagoner; Animesh Tandon; Mina K. Chung; Anant Madabhushi

doi:10.1161/CIRCEP.123.012679

Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation: A Multi-Site Study

Golnoush Asaeikheybari, Majd El-Harasis, Amit Gupta, M. Benjamin Shoemaker, John Barnard, Joshua Hunter, Rod S. Passman, Han Sun, Hyun Su Kim, Taylor Schilling, William Telfer, Britta Eldridge, Po Hao Chen, Abhishek Midya, Bibin Varghese, Samuel J. Harwood, Alison Jin, Sojin Y. Wass, Aleksandar Izda, Kevin ParkAbel Abraham, David R. Van Wagoner, Animesh Tandon, Mina K. Chung^*, Anant Madabhushi^*

^*Corresponding author for this work

Medicine, Cardiology Division

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

Abstract

BACKGROUND: Atrial fibrillation (AF) recurrence is common after catheter ablation. Pulmonary vein (PV) isolation is the cornerstone of AF ablation, but PV remodeling has been associated with the risk of AF recurrence. We aimed to evaluate whether artificial intelligence-based morphological features of primary and secondary PV branches on computed tomography images are associated with AF recurrence post-ablation. METHODS: Two artificial intelligence models were trained for the segmentation of computed tomography images, enabling the isolation of PV branches. Patients from Cleveland Clinic (N=135) and Vanderbilt University (N=594) were combined and divided into 2 sets for training and cross-validation (D₁, n=218) and internal testing (D₂, n=511). An independent validation set (D₃, N=80) was obtained from University Hospitals of Cleveland. We extracted 48 fractal-based and 12 shape-based radiomic features from primary and secondary PV branches of patients with AF recurrence (AF+) and without recurrence after catheter ablation of AF (AF-). To predict AFrecurrence, 3 Gradient Boosting classification models based on significant features from primary (M_p), secondary (M_s), and combined (M_c) PV branches were built. RESULTS: Features relating to primary PVs were found to be associated with AF recurrence. The M_p classifier achieved area under the curve values of 0.73, 0.71, and 0.70 across the 3 datasets. AF+ cases exhibited greater surface complexity in their primary PV area, as evidenced by higher fractal dimension values compared with AF- cases. The M_s classifier results revealed a weaker association with AF+, suggesting higher relevance to AF recurrence post-ablation from primary PV branch morphology. CONCLUSIONS: This largest multi-institutional study to date revealed associations between artificial intelligence-extracted morphological features of the primary PV branches with AF recurrence in 809 patients from 3 sites. Future work will focus on enhancing the predictive ability of the classifier by integrating clinical, structural, and morphological features, including left atrial appendage and left atrium-related characteristics.

Original language	English (US)
Pages (from-to)	e012679
Journal	Circulation: Arrhythmia and Electrophysiology
Volume	17
Issue number	12
DOIs	https://doi.org/10.1161/CIRCEP.123.012679
State	Published - Dec 1 2024

Funding

For Drs Chung, Barnard, Van Wagoner, KS, and H.S. Kim were supported by National Institutes of Health (NIH) grants R01 HL 111314, R01 HL 090620, P01HL158502; American Heart Association Atrial Fibrillation Strategically Focused Research Network grants 18SFRN34110067, 20SCG35490449, 18SFRN34170013, 18SFRN34170442; the NIH National Center for Research Resources for Case Western Reserve University and Cleveland Clinic Clinical and Translational Science Award UL1-RR024989; the Cleveland Clinic Department of Cardiovascular Medicine philanthropy research funds; and the Tomsich Atrial Fibrillation Research Fund . Dr Tandon was supported in part by NIH/NHLBI K23HL150279.

Keywords

artificial intelligence
atrial fibrillation
catheter ablation
fractals
pulmonary veins

ASJC Scopus subject areas

Cardiology and Cardiovascular Medicine
Physiology (medical)

Access to Document

10.1161/CIRCEP.123.012679

Cite this

Asaeikheybari, G., El-Harasis, M., Gupta, A., Shoemaker, M. B., Barnard, J., Hunter, J., Passman, R. S., Sun, H., Kim, H. S., Schilling, T., Telfer, W., Eldridge, B., Chen, P. H., Midya, A., Varghese, B., Harwood, S. J., Jin, A., Wass, S. Y., Izda, A., ... Madabhushi, A. (2024). Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation: A Multi-Site Study. Circulation: Arrhythmia and Electrophysiology, 17(12), e012679. https://doi.org/10.1161/CIRCEP.123.012679

@article{0163028153184802975128c53150a85a,

title = "Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation: A Multi-Site Study",

abstract = "BACKGROUND: Atrial fibrillation (AF) recurrence is common after catheter ablation. Pulmonary vein (PV) isolation is the cornerstone of AF ablation, but PV remodeling has been associated with the risk of AF recurrence. We aimed to evaluate whether artificial intelligence-based morphological features of primary and secondary PV branches on computed tomography images are associated with AF recurrence post-ablation. METHODS: Two artificial intelligence models were trained for the segmentation of computed tomography images, enabling the isolation of PV branches. Patients from Cleveland Clinic (N=135) and Vanderbilt University (N=594) were combined and divided into 2 sets for training and cross-validation (D1, n=218) and internal testing (D2, n=511). An independent validation set (D3, N=80) was obtained from University Hospitals of Cleveland. We extracted 48 fractal-based and 12 shape-based radiomic features from primary and secondary PV branches of patients with AF recurrence (AF+) and without recurrence after catheter ablation of AF (AF-). To predict AFrecurrence, 3 Gradient Boosting classification models based on significant features from primary (Mp), secondary (Ms), and combined (Mc) PV branches were built. RESULTS: Features relating to primary PVs were found to be associated with AF recurrence. The Mp classifier achieved area under the curve values of 0.73, 0.71, and 0.70 across the 3 datasets. AF+ cases exhibited greater surface complexity in their primary PV area, as evidenced by higher fractal dimension values compared with AF- cases. The Ms classifier results revealed a weaker association with AF+, suggesting higher relevance to AF recurrence post-ablation from primary PV branch morphology. CONCLUSIONS: This largest multi-institutional study to date revealed associations between artificial intelligence-extracted morphological features of the primary PV branches with AF recurrence in 809 patients from 3 sites. Future work will focus on enhancing the predictive ability of the classifier by integrating clinical, structural, and morphological features, including left atrial appendage and left atrium-related characteristics.",

keywords = "artificial intelligence, atrial fibrillation, catheter ablation, fractals, pulmonary veins",

author = "Golnoush Asaeikheybari and Majd El-Harasis and Amit Gupta and Shoemaker, {M. Benjamin} and John Barnard and Joshua Hunter and Passman, {Rod S.} and Han Sun and Kim, {Hyun Su} and Taylor Schilling and William Telfer and Britta Eldridge and Chen, {Po Hao} and Abhishek Midya and Bibin Varghese and Harwood, {Samuel J.} and Alison Jin and Wass, {Sojin Y.} and Aleksandar Izda and Kevin Park and Abel Abraham and {Van Wagoner}, {David R.} and Animesh Tandon and Chung, {Mina K.} and Anant Madabhushi",

note = "Publisher Copyright: {\textcopyright} 2024 The Authors.",

year = "2024",

month = dec,

day = "1",

doi = "10.1161/CIRCEP.123.012679",

language = "English (US)",

volume = "17",

pages = "e012679",

journal = "Circulation: Arrhythmia and Electrophysiology",

issn = "1941-3149",

publisher = "Lippincott Williams and Wilkins",

number = "12",

}

Asaeikheybari, G, El-Harasis, M, Gupta, A, Shoemaker, MB, Barnard, J, Hunter, J, Passman, RS, Sun, H, Kim, HS, Schilling, T, Telfer, W, Eldridge, B, Chen, PH, Midya, A, Varghese, B, Harwood, SJ, Jin, A, Wass, SY, Izda, A, Park, K, Abraham, A, Van Wagoner, DR, Tandon, A, Chung, MK & Madabhushi, A 2024, 'Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation: A Multi-Site Study', Circulation: Arrhythmia and Electrophysiology, vol. 17, no. 12, pp. e012679. https://doi.org/10.1161/CIRCEP.123.012679

Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation: A Multi-Site Study. / Asaeikheybari, Golnoush; El-Harasis, Majd; Gupta, Amit et al.
In: Circulation: Arrhythmia and Electrophysiology, Vol. 17, No. 12, 01.12.2024, p. e012679.

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

TY - JOUR

T1 - Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation

T2 - A Multi-Site Study

AU - Asaeikheybari, Golnoush

AU - El-Harasis, Majd

AU - Gupta, Amit

AU - Shoemaker, M. Benjamin

AU - Barnard, John

AU - Hunter, Joshua

AU - Passman, Rod S.

AU - Sun, Han

AU - Kim, Hyun Su

AU - Schilling, Taylor

AU - Telfer, William

AU - Eldridge, Britta

AU - Chen, Po Hao

AU - Midya, Abhishek

AU - Varghese, Bibin

AU - Harwood, Samuel J.

AU - Jin, Alison

AU - Wass, Sojin Y.

AU - Izda, Aleksandar

AU - Park, Kevin

AU - Abraham, Abel

AU - Van Wagoner, David R.

AU - Tandon, Animesh

AU - Chung, Mina K.

AU - Madabhushi, Anant

PY - 2024/12/1

Y1 - 2024/12/1

N2 - BACKGROUND: Atrial fibrillation (AF) recurrence is common after catheter ablation. Pulmonary vein (PV) isolation is the cornerstone of AF ablation, but PV remodeling has been associated with the risk of AF recurrence. We aimed to evaluate whether artificial intelligence-based morphological features of primary and secondary PV branches on computed tomography images are associated with AF recurrence post-ablation. METHODS: Two artificial intelligence models were trained for the segmentation of computed tomography images, enabling the isolation of PV branches. Patients from Cleveland Clinic (N=135) and Vanderbilt University (N=594) were combined and divided into 2 sets for training and cross-validation (D1, n=218) and internal testing (D2, n=511). An independent validation set (D3, N=80) was obtained from University Hospitals of Cleveland. We extracted 48 fractal-based and 12 shape-based radiomic features from primary and secondary PV branches of patients with AF recurrence (AF+) and without recurrence after catheter ablation of AF (AF-). To predict AFrecurrence, 3 Gradient Boosting classification models based on significant features from primary (Mp), secondary (Ms), and combined (Mc) PV branches were built. RESULTS: Features relating to primary PVs were found to be associated with AF recurrence. The Mp classifier achieved area under the curve values of 0.73, 0.71, and 0.70 across the 3 datasets. AF+ cases exhibited greater surface complexity in their primary PV area, as evidenced by higher fractal dimension values compared with AF- cases. The Ms classifier results revealed a weaker association with AF+, suggesting higher relevance to AF recurrence post-ablation from primary PV branch morphology. CONCLUSIONS: This largest multi-institutional study to date revealed associations between artificial intelligence-extracted morphological features of the primary PV branches with AF recurrence in 809 patients from 3 sites. Future work will focus on enhancing the predictive ability of the classifier by integrating clinical, structural, and morphological features, including left atrial appendage and left atrium-related characteristics.

AB - BACKGROUND: Atrial fibrillation (AF) recurrence is common after catheter ablation. Pulmonary vein (PV) isolation is the cornerstone of AF ablation, but PV remodeling has been associated with the risk of AF recurrence. We aimed to evaluate whether artificial intelligence-based morphological features of primary and secondary PV branches on computed tomography images are associated with AF recurrence post-ablation. METHODS: Two artificial intelligence models were trained for the segmentation of computed tomography images, enabling the isolation of PV branches. Patients from Cleveland Clinic (N=135) and Vanderbilt University (N=594) were combined and divided into 2 sets for training and cross-validation (D1, n=218) and internal testing (D2, n=511). An independent validation set (D3, N=80) was obtained from University Hospitals of Cleveland. We extracted 48 fractal-based and 12 shape-based radiomic features from primary and secondary PV branches of patients with AF recurrence (AF+) and without recurrence after catheter ablation of AF (AF-). To predict AFrecurrence, 3 Gradient Boosting classification models based on significant features from primary (Mp), secondary (Ms), and combined (Mc) PV branches were built. RESULTS: Features relating to primary PVs were found to be associated with AF recurrence. The Mp classifier achieved area under the curve values of 0.73, 0.71, and 0.70 across the 3 datasets. AF+ cases exhibited greater surface complexity in their primary PV area, as evidenced by higher fractal dimension values compared with AF- cases. The Ms classifier results revealed a weaker association with AF+, suggesting higher relevance to AF recurrence post-ablation from primary PV branch morphology. CONCLUSIONS: This largest multi-institutional study to date revealed associations between artificial intelligence-extracted morphological features of the primary PV branches with AF recurrence in 809 patients from 3 sites. Future work will focus on enhancing the predictive ability of the classifier by integrating clinical, structural, and morphological features, including left atrial appendage and left atrium-related characteristics.

KW - artificial intelligence

KW - atrial fibrillation

KW - catheter ablation

KW - fractals

KW - pulmonary veins

UR - http://www.scopus.com/inward/record.url?scp=85211597147&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85211597147&partnerID=8YFLogxK

U2 - 10.1161/CIRCEP.123.012679

DO - 10.1161/CIRCEP.123.012679

M3 - Article

C2 - 39624901

AN - SCOPUS:85211597147

SN - 1941-3149

VL - 17

SP - e012679

JO - Circulation: Arrhythmia and Electrophysiology

JF - Circulation: Arrhythmia and Electrophysiology

IS - 12

ER -

Asaeikheybari G, El-Harasis M, Gupta A, Shoemaker MB, Barnard J, Hunter J et al. Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation: A Multi-Site Study. Circulation: Arrhythmia and Electrophysiology. 2024 Dec 1;17(12):e012679. doi: 10.1161/CIRCEP.123.012679

Artificial Intelligence-Based Feature Analysis of Pulmonary Vein Morphology on Computed Tomography Scans and Risk of Atrial Fibrillation Recurrence after Catheter Ablation: A Multi-Site Study

Abstract

Funding

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this