Patient-specific myocardial infarction risk thresholds from AI-enabled coronary plaque analysis

Robert JH Miller; Nipun  Manral; Andrew  Lin; Aakash  Shanbhag; Caroline  Park; Jacek Kwiecinski; Aditya  Killekar; Priscilla  McElhinney; Hidenari  Matsumoto; Aryabod  Razipour; Kajetan  Grodecki; Alan C. Kwan; Donghee Han; Keiichiro  Kuronuma; Guadalupe Flores  Tomasino; Jolien Geers; Markus Goeller; Mohamed  Marwan; Heidi Gransar; Balaji K Tamarappoo; Sebastian Cadet; Victor Y Cheng; Stephen J Nicholls; Dennis T  Wong; Lu Chen; Jane Cao; Daniel S Berman; Marc R Dweck; David E Newby; Michelle C Williams; Piotr Slomka; Damini Dey

doi:10.1161/CIRCIMAGING.124.016958

Patient-specific myocardial infarction risk thresholds from AI-enabled coronary plaque analysis

Robert JH Miller, Nipun Manral, Andrew Lin, Aakash Shanbhag, Caroline Park, Jacek Kwiecinski, Aditya Killekar, Priscilla McElhinney , Hidenari Matsumoto, Aryabod Razipour, Kajetan Grodecki, Alan C. Kwan, Donghee Han, Keiichiro Kuronuma, Guadalupe Flores Tomasino, Jolien Geers, Markus Goeller, Mohamed Marwan, Heidi Gransar, Balaji K TamarappooSebastian Cadet, Victor Y Cheng, Stephen J Nicholls, Dennis T Wong, Lu Chen, Jane Cao, Daniel S Berman, Marc R Dweck, David E Newby, Michelle C Williams, Piotr Slomka, Damini Dey

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

Abstract

Background: Plaque quantification from coronary computed tomography angiography (CTA) has emerged as a valuable predictor of cardiovascular risk. Deep learning (DL) can provide automated quantification of coronary plaque from CTA. We determined per-patient age and sex-specific distributions of DL-based plaque measurements and further evaluated their risk prediction for myocardial infarction in external samples.
Methods: In this international, multicenter study of 2803 patients, a previously validated DL system was used to quantify coronary plaque from CTA. Age and sex-specific distributions of coronary plaque volume were determined from 956 patients undergoing CTA for stable coronary artery disease from 5 cohorts. Multicenter external samples were used to evaluate associations between coronary plaque percentiles and myocardial infarction.
Results: Quantitative DL plaque volumes increased with age and were higher in male patients. In the combined external sample (n=1,847), patients in the ≥75th percentile of total plaque volume (unadjusted hazard ratio 2.65, 95% confidence interval 1.47-4.78, p=0.001) were at increased risk of myocardial infarction compared to patients below the 50th percentile. Similar relationships were seen for most plaque volumes and persisted in multivariable analyses adjusting for clinical characteristics, coronary artery calcium, stenosis and plaque volume, with adjusted hazard ratios ranging from 2.38 to 2.50 for patients in the ≥75th percentile of total plaque volume.
Conclusions: Per-patient age and sex-specific distributions for deep learning-based coronary plaque volumes are strongly predictive of myocardial infarction, with the highest risk seen in patients with coronary plaque volumes in the ≥75th percentile.
Keywords: Deep learning; coronary plaque; risk prediction; coronary CT Angiography; sex-specific analysis; myocardial infarction

Original language	English
Journal	Circulation: Cardiovascular Imaging
Early online date	30 Sept 2024
DOIs	https://doi.org/10.1161/CIRCIMAGING.124.016958
Publication status	E-pub ahead of print - 30 Sept 2024

Access to Document

10.1161/CIRCIMAGING.124.016958

Manuscript Circ CVI 6-27 CleanAccepted author manuscript, 1.7 MBLicence: Creative Commons: Attribution (CC-BY)

Cite this

Miller, R. JH., Manral, N., Lin, A., Shanbhag, A., Park, C., Kwiecinski, J., Killekar, A., McElhinney , P., Matsumoto, H., Razipour, A., Grodecki, K., Kwan, A. C., Han, D., Kuronuma, K., Tomasino, G. F., Geers, J., Goeller, M., Marwan, M., Gransar, H., ... Dey, D. (2024). Patient-specific myocardial infarction risk thresholds from AI-enabled coronary plaque analysis. Circulation: Cardiovascular Imaging. Advance online publication. https://doi.org/10.1161/CIRCIMAGING.124.016958

@article{c18fc2f186834a7fbdacae21e484efc5,

title = "Patient-specific myocardial infarction risk thresholds from AI-enabled coronary plaque analysis",

abstract = "Background: Plaque quantification from coronary computed tomography angiography (CTA) has emerged as a valuable predictor of cardiovascular risk. Deep learning (DL) can provide automated quantification of coronary plaque from CTA. We determined per-patient age and sex-specific distributions of DL-based plaque measurements and further evaluated their risk prediction for myocardial infarction in external samples.Methods: In this international, multicenter study of 2803 patients, a previously validated DL system was used to quantify coronary plaque from CTA. Age and sex-specific distributions of coronary plaque volume were determined from 956 patients undergoing CTA for stable coronary artery disease from 5 cohorts. Multicenter external samples were used to evaluate associations between coronary plaque percentiles and myocardial infarction.Results: Quantitative DL plaque volumes increased with age and were higher in male patients. In the combined external sample (n=1,847), patients in the ≥75th percentile of total plaque volume (unadjusted hazard ratio 2.65, 95% confidence interval 1.47-4.78, p=0.001) were at increased risk of myocardial infarction compared to patients below the 50th percentile. Similar relationships were seen for most plaque volumes and persisted in multivariable analyses adjusting for clinical characteristics, coronary artery calcium, stenosis and plaque volume, with adjusted hazard ratios ranging from 2.38 to 2.50 for patients in the ≥75th percentile of total plaque volume. Conclusions: Per-patient age and sex-specific distributions for deep learning-based coronary plaque volumes are strongly predictive of myocardial infarction, with the highest risk seen in patients with coronary plaque volumes in the ≥75th percentile.Keywords: Deep learning; coronary plaque; risk prediction; coronary CT Angiography; sex-specific analysis; myocardial infarction",

author = "Miller, {Robert JH} and Nipun Manral and Andrew Lin and Aakash Shanbhag and Caroline Park and Jacek Kwiecinski and Aditya Killekar and Priscilla McElhinney and Hidenari Matsumoto and Aryabod Razipour and Kajetan Grodecki and Kwan, {Alan C.} and Donghee Han and Keiichiro Kuronuma and Tomasino, {Guadalupe Flores} and Jolien Geers and Markus Goeller and Mohamed Marwan and Heidi Gransar and Tamarappoo, {Balaji K} and Sebastian Cadet and Cheng, {Victor Y} and Nicholls, {Stephen J} and Wong, {Dennis T} and Lu Chen and Jane Cao and Berman, {Daniel S} and Dweck, {Marc R} and Newby, {David E} and Williams, {Michelle C} and Piotr Slomka and Damini Dey",

year = "2024",

month = sep,

day = "30",

doi = "10.1161/CIRCIMAGING.124.016958",

language = "English",

journal = "Circulation: Cardiovascular Imaging",

issn = "1941-9651",

publisher = "Lippincott Williams and Wilkins Ltd.",

}

Miller, RJH, Manral, N, Lin, A, Shanbhag, A, Park, C, Kwiecinski, J, Killekar, A, McElhinney , P, Matsumoto, H, Razipour, A, Grodecki, K, Kwan, AC, Han, D, Kuronuma, K, Tomasino, GF, Geers, J, Goeller, M, Marwan, M, Gransar, H, Tamarappoo, BK, Cadet, S, Cheng, VY, Nicholls, SJ, Wong, DT, Chen, L, Cao, J, Berman, DS, Dweck, MR , Newby, DE , Williams, MC, Slomka, P & Dey, D 2024, 'Patient-specific myocardial infarction risk thresholds from AI-enabled coronary plaque analysis', Circulation: Cardiovascular Imaging. https://doi.org/10.1161/CIRCIMAGING.124.016958

TY - JOUR

T1 - Patient-specific myocardial infarction risk thresholds from AI-enabled coronary plaque analysis

AU - Miller, Robert JH

AU - Manral, Nipun

AU - Lin, Andrew

AU - Shanbhag, Aakash

AU - Park, Caroline

AU - Kwiecinski, Jacek

AU - Killekar, Aditya

AU - McElhinney , Priscilla

AU - Matsumoto, Hidenari

AU - Razipour, Aryabod

AU - Grodecki, Kajetan

AU - Kwan, Alan C.

AU - Han, Donghee

AU - Kuronuma, Keiichiro

AU - Tomasino, Guadalupe Flores

AU - Geers, Jolien

AU - Goeller, Markus

AU - Marwan, Mohamed

AU - Gransar, Heidi

AU - Tamarappoo, Balaji K

AU - Cadet, Sebastian

AU - Cheng, Victor Y

AU - Nicholls, Stephen J

AU - Wong, Dennis T

AU - Chen, Lu

AU - Cao, Jane

AU - Berman, Daniel S

AU - Dweck, Marc R

AU - Newby, David E

AU - Williams, Michelle C

AU - Slomka, Piotr

AU - Dey, Damini

PY - 2024/9/30

Y1 - 2024/9/30

N2 - Background: Plaque quantification from coronary computed tomography angiography (CTA) has emerged as a valuable predictor of cardiovascular risk. Deep learning (DL) can provide automated quantification of coronary plaque from CTA. We determined per-patient age and sex-specific distributions of DL-based plaque measurements and further evaluated their risk prediction for myocardial infarction in external samples.Methods: In this international, multicenter study of 2803 patients, a previously validated DL system was used to quantify coronary plaque from CTA. Age and sex-specific distributions of coronary plaque volume were determined from 956 patients undergoing CTA for stable coronary artery disease from 5 cohorts. Multicenter external samples were used to evaluate associations between coronary plaque percentiles and myocardial infarction.Results: Quantitative DL plaque volumes increased with age and were higher in male patients. In the combined external sample (n=1,847), patients in the ≥75th percentile of total plaque volume (unadjusted hazard ratio 2.65, 95% confidence interval 1.47-4.78, p=0.001) were at increased risk of myocardial infarction compared to patients below the 50th percentile. Similar relationships were seen for most plaque volumes and persisted in multivariable analyses adjusting for clinical characteristics, coronary artery calcium, stenosis and plaque volume, with adjusted hazard ratios ranging from 2.38 to 2.50 for patients in the ≥75th percentile of total plaque volume. Conclusions: Per-patient age and sex-specific distributions for deep learning-based coronary plaque volumes are strongly predictive of myocardial infarction, with the highest risk seen in patients with coronary plaque volumes in the ≥75th percentile.Keywords: Deep learning; coronary plaque; risk prediction; coronary CT Angiography; sex-specific analysis; myocardial infarction

AB - Background: Plaque quantification from coronary computed tomography angiography (CTA) has emerged as a valuable predictor of cardiovascular risk. Deep learning (DL) can provide automated quantification of coronary plaque from CTA. We determined per-patient age and sex-specific distributions of DL-based plaque measurements and further evaluated their risk prediction for myocardial infarction in external samples.Methods: In this international, multicenter study of 2803 patients, a previously validated DL system was used to quantify coronary plaque from CTA. Age and sex-specific distributions of coronary plaque volume were determined from 956 patients undergoing CTA for stable coronary artery disease from 5 cohorts. Multicenter external samples were used to evaluate associations between coronary plaque percentiles and myocardial infarction.Results: Quantitative DL plaque volumes increased with age and were higher in male patients. In the combined external sample (n=1,847), patients in the ≥75th percentile of total plaque volume (unadjusted hazard ratio 2.65, 95% confidence interval 1.47-4.78, p=0.001) were at increased risk of myocardial infarction compared to patients below the 50th percentile. Similar relationships were seen for most plaque volumes and persisted in multivariable analyses adjusting for clinical characteristics, coronary artery calcium, stenosis and plaque volume, with adjusted hazard ratios ranging from 2.38 to 2.50 for patients in the ≥75th percentile of total plaque volume. Conclusions: Per-patient age and sex-specific distributions for deep learning-based coronary plaque volumes are strongly predictive of myocardial infarction, with the highest risk seen in patients with coronary plaque volumes in the ≥75th percentile.Keywords: Deep learning; coronary plaque; risk prediction; coronary CT Angiography; sex-specific analysis; myocardial infarction

U2 - 10.1161/CIRCIMAGING.124.016958

DO - 10.1161/CIRCIMAGING.124.016958

M3 - Article

SN - 1941-9651

JO - Circulation: Cardiovascular Imaging

JF - Circulation: Cardiovascular Imaging

ER -

Patient-specific myocardial infarction risk thresholds from AI-enabled coronary plaque analysis

Abstract

Access to Document

Fingerprint

Cite this