Development and validation of AI-derived segmentation of four-chamber cine cardiac magnetic resonance.
Artificial intelligence, Deep learning, Heart diseases, Magnetic resonance imaging (cine), Prognosis
Journal
European radiology experimental
ISSN: 2509-9280
Titre abrégé: Eur Radiol Exp
Pays: England
ID NLM: 101721752
Informations de publication
Date de publication:
12 Jul 2024
12 Jul 2024
Historique:
received:
13
02
2024
accepted:
30
04
2024
medline:
12
7
2024
pubmed:
12
7
2024
entrez:
11
7
2024
Statut:
epublish
Résumé
Cardiac magnetic resonance (CMR) in the four-chamber plane offers comprehensive insight into the volumetrics of the heart. We aimed to develop an artificial intelligence (AI) model of time-resolved segmentation using the four-chamber cine. A fully automated deep learning algorithm was trained using retrospective multicentre and multivendor data of 814 subjects. Validation, reproducibility, and mortality prediction were evaluated on an independent cohort of 101 subjects. The mean age of the validation cohort was 54 years, and 66 (65%) were males. Left and right heart parameters demonstrated strong correlations between automated and manual analysis, with a ρ of 0.91-0.98 and 0.89-0.98, respectively, with minimal bias. All AI four-chamber volumetrics in repeatability analysis demonstrated high correlation (ρ = 0.99-1.00) and no bias. Automated four-chamber analysis underestimated both left ventricular (LV) and right ventricular (RV) volumes compared to ground-truth short-axis cine analysis. Two correction factors for LV and RV four-chamber analysis were proposed based on systematic bias. After applying the correction factors, a strong correlation and minimal bias for LV volumetrics were observed. During a mean follow-up period of 6.75 years, 16 patients died. On stepwise multivariable analysis, left atrial ejection fraction demonstrated an independent association with death in both manual (hazard ratio (HR) = 0.96, p = 0.003) and AI analyses (HR = 0.96, p < 0.001). Fully automated four-chamber CMR is feasible, reproducible, and has the same real-world prognostic value as manual analysis. LV volumes by four-chamber segmentation were comparable to short-axis volumetric assessment. ClinicalTrials.gov: NCT05114785. Integrating fully automated AI in CMR promises to revolutionise clinical cardiac assessment, offering efficient, accurate, and prognostically valuable insights for improved patient care and outcomes. • Four-chamber cine sequences remain one of the most informative acquisitions in CMR examination. • This deep learning-based, time-resolved, fully automated four-chamber volumetric, functional, and deformation analysis solution. • LV and RV were underestimated by four-chamber analysis compared to ground truth short-axis segmentation. • Correction bias for both LV and RV volumes by four-chamber segmentation, minimises the systematic bias.
Sections du résumé
BACKGROUND
BACKGROUND
Cardiac magnetic resonance (CMR) in the four-chamber plane offers comprehensive insight into the volumetrics of the heart. We aimed to develop an artificial intelligence (AI) model of time-resolved segmentation using the four-chamber cine.
METHODS
METHODS
A fully automated deep learning algorithm was trained using retrospective multicentre and multivendor data of 814 subjects. Validation, reproducibility, and mortality prediction were evaluated on an independent cohort of 101 subjects.
RESULTS
RESULTS
The mean age of the validation cohort was 54 years, and 66 (65%) were males. Left and right heart parameters demonstrated strong correlations between automated and manual analysis, with a ρ of 0.91-0.98 and 0.89-0.98, respectively, with minimal bias. All AI four-chamber volumetrics in repeatability analysis demonstrated high correlation (ρ = 0.99-1.00) and no bias. Automated four-chamber analysis underestimated both left ventricular (LV) and right ventricular (RV) volumes compared to ground-truth short-axis cine analysis. Two correction factors for LV and RV four-chamber analysis were proposed based on systematic bias. After applying the correction factors, a strong correlation and minimal bias for LV volumetrics were observed. During a mean follow-up period of 6.75 years, 16 patients died. On stepwise multivariable analysis, left atrial ejection fraction demonstrated an independent association with death in both manual (hazard ratio (HR) = 0.96, p = 0.003) and AI analyses (HR = 0.96, p < 0.001).
CONCLUSION
CONCLUSIONS
Fully automated four-chamber CMR is feasible, reproducible, and has the same real-world prognostic value as manual analysis. LV volumes by four-chamber segmentation were comparable to short-axis volumetric assessment.
TRIALS REGISTRATION
BACKGROUND
ClinicalTrials.gov: NCT05114785.
RELEVANCE STATEMENT
CONCLUSIONS
Integrating fully automated AI in CMR promises to revolutionise clinical cardiac assessment, offering efficient, accurate, and prognostically valuable insights for improved patient care and outcomes.
KEY POINTS
CONCLUSIONS
• Four-chamber cine sequences remain one of the most informative acquisitions in CMR examination. • This deep learning-based, time-resolved, fully automated four-chamber volumetric, functional, and deformation analysis solution. • LV and RV were underestimated by four-chamber analysis compared to ground truth short-axis segmentation. • Correction bias for both LV and RV volumes by four-chamber segmentation, minimises the systematic bias.
Identifiants
pubmed: 38992116
doi: 10.1186/s41747-024-00477-7
pii: 10.1186/s41747-024-00477-7
doi:
Banques de données
ClinicalTrials.gov
['NCT05114785']
Types de publication
Journal Article
Validation Study
Multicenter Study
Langues
eng
Sous-ensembles de citation
IM
Pagination
77Subventions
Organisme : Wellcome Trust
ID : 220703/Z/20/Z
Pays : United Kingdom
Informations de copyright
© 2024. The Author(s).
Références
Leiner T, Bogaert J, Friedrich MG et al (2020) SCMR Position Paper (2020) on clinical indications for cardiovascular magnetic resonance. J Cardiovasc Magn Reson 22:76. https://doi.org/10.1186/s12968-020-00682-4
doi: 10.1186/s12968-020-00682-4
pubmed: 33161900
pmcid: 7649060
Mariscal-Harana J, Asher C, Vergani V et al (2023) An artificial intelligence tool for automated analysis of large-scale unstructured clinical cine cardiac magnetic resonance databases. Eur Heart J Digit Health 4:370–383. https://doi.org/10.1093/ehjdh/ztad044
doi: 10.1093/ehjdh/ztad044
pubmed: 37794871
pmcid: 10545512
Bai W, Sinclair M, Tarroni G et al (2018) Automated cardiovascular magnetic resonance image analysis with fully convolutional networks. J Cardiovasc Magn Reson 20:1–12. https://doi.org/10.1186/s12968-018-0471-x
doi: 10.1186/s12968-018-0471-x
Alabed S, Maiter A, Mahmood A et al (2022) Quality of reporting of artificial intelligence studies: Lessons learnt from a systematic review of the literature. Clin Radiol 77:e21. https://doi.org/10.1016/J.CRAD.2022.09.057
doi: 10.1016/J.CRAD.2022.09.057
Suinesiaputra A, Mauger CA, Ambale-Venkatesh B et al (2022) Deep Learning Analysis of Cardiac MRI in Legacy Datasets: Multi-Ethnic Study of Atherosclerosis. Front Cardiovasc Med 8:807728. https://doi.org/10.3389/FCVM.2021.807728/BIBTEX
doi: 10.3389/FCVM.2021.807728/BIBTEX
pubmed: 35127868
pmcid: 8813768
Penso M, Moccia S, Scafuri S, et al (2021) Automated left and right ventricular chamber segmentation in cardiac magnetic resonance images using dense fully convolutional neural network. Comput Methods Programs Biomed 204. https://doi.org/10.1016/J.CMPB.2021.106059
Fadil H, Totman JJ, Hausenloy DJ et al (2021) A deep learning pipeline for automatic analysis of multi-scan cardiovascular magnetic resonance. J Cardiovasc Magn Reson 23:1–13. https://doi.org/10.1186/S12968-020-00695-Z/FIGURES/9
doi: 10.1186/S12968-020-00695-Z/FIGURES/9
Alabed S, Alandejani F, Dwivedi K et al (2022) Validation of Artificial Intelligence Cardiac MRI Measurements: Relationship to Heart Catheterization and Mortality Prediction. Radiology 304:68–79. https://doi.org/10.1148/RADIOL.229014
doi: 10.1148/RADIOL.229014
Kramer CM, Barkhausen J, Bucciarelli-Ducci C, et al (2020) Standardized cardiovascular magnetic resonance imaging (CMR) protocols: 2020 update. J Cardiovasc Magn Reson 22. https://doi.org/10.1186/S12968-020-00607-1
Schulz-Menger J, Bluemke DA, Bremerich J, et al (2013) Standardized image interpretation and post processing in cardiovascular magnetic resonance: Society for Cardiovascular Magnetic Resonance (SCMR) Board of Trustees Task Force on Standardized Post Processing. J Cardiovasc Magn Reson 15:. https://doi.org/10.1186/1532-429X-15-35
Schulz-Menger J, Bluemke DA, Bremerich J et al (2020) Standardized image interpretation and post-processing in cardiovascular magnetic resonance - 2020 update: Society for Cardiovascular Magnetic Resonance (SCMR): Board of Trustees Task Force on Standardized Post-Processing. J Cardiovasc Magn Reson 22:1–22. https://doi.org/10.1186/S12968-020-00610-6/FIGURES/10
doi: 10.1186/S12968-020-00610-6/FIGURES/10
Shahzad R, Tao Q, Dzyubachyk O et al (2017) Fully-automatic left ventricular segmentation from long-axis cardiac cine MR scans. Med Image Anal 39:44–55. https://doi.org/10.1016/J.MEDIA.2017.04.004
doi: 10.1016/J.MEDIA.2017.04.004
pubmed: 28432954
Gonzales RA, Seemann F, Lamy J, et al (2021) Automated left atrial time-resolved segmentation in MRI long-axis cine images using active contours. BMC Med Imaging 21. https://doi.org/10.1186/S12880-021-00630-3
Guensch DP, Kuganathan S, Utz CD et al (2023) Analysis of bi-atrial function using CMR feature tracking and long-axis shortening approaches in patients with diastolic dysfunction and atrial fibrillation. Eur Radiol 33:7226–7237. https://doi.org/10.1007/S00330-023-09663-4/TABLES/1
doi: 10.1007/S00330-023-09663-4/TABLES/1
pubmed: 37145149
pmcid: 10511591
Alandejani F, Alabed S, Garg P, et al (2022) Training and clinical testing of artificial intelligence derived right atrial cardiovascular magnetic resonance measurements. J Cardiovasc Magn Reson 24:. https://doi.org/10.1186/S12968-022-00855-3
Garg P, Crandon S, Swoboda PP et al (2018) Left ventricular blood flow kinetic energy after myocardial infarction - Insights from 4D flow cardiovascular magnetic resonance. J Cardiovasc Magn Reson 20:1–15. https://doi.org/10.1186/S12968-018-0483-6/FIGURES/8
doi: 10.1186/S12968-018-0483-6/FIGURES/8
Crandon S, Westenberg JJM, Swoboda PP et al (2018) Impact of Age and Diastolic Function on Novel, 4D flow CMR Biomarkers of Left Ventricular Blood Flow Kinetic Energy. Sci Rep 8(18):1–11. https://doi.org/10.1038/s41598-018-32707-5
doi: 10.1038/s41598-018-32707-5
Cardillo L, Cahill F, Wylie H et al (2018) Patients’ perspectives on opt-out consent for observational research: systematic review and focus group. Br J Nurs 27:1321–29. https://doi.org/10.12968/BJON.2018.27.22.1321
doi: 10.12968/BJON.2018.27.22.1321
pubmed: 30525978
Vellinga A, Cormican M, Hanahoe B, et al (2011) Opt-out as an acceptable method of obtaining consent in medical research: a short report. BMC Med Res Methodol 11. https://doi.org/10.1186/1471-2288-11-40
Papavassiliu T, Kühl HP, Schröder M et al (2005) Effect of endocardial trabeculae on left ventricular measurements and measurement reproducibility at cardiovascular MR imaging. Radiology 236:57–64. https://doi.org/10.1148/RADIOL.2353040601
doi: 10.1148/RADIOL.2353040601
pubmed: 15955850
Hurdman J, Condliffe R, Elliot CA et al (2012) ASPIRE registry: Assessing the Spectrum of Pulmonary hypertension Identified at a REferral centre. Eur Respir J 39:945–955. https://doi.org/10.1183/09031936.00078411
doi: 10.1183/09031936.00078411
pubmed: 21885399
Chew PG, Dobson LE, Garg P et al (2019) CMR quantitation of change in mitral regurgitation following transcatheter aortic valve replacement (TAVR): impact on left ventricular reverse remodeling and outcome. Int J Cardiovasc Imaging 35:161–170. https://doi.org/10.1007/S10554-018-1441-Y/FIGURES/4
doi: 10.1007/S10554-018-1441-Y/FIGURES/4
pubmed: 30182320
Ruijsink B, Puyol-Antón E, Oksuz I et al (2020) Fully Automated, Quality-Controlled Cardiac Analysis From CMR: Validation and Large-Scale Application to Characterize Cardiac Function. JACC Cardiovasc Imaging 13:684–695. https://doi.org/10.1016/J.JCMG.2019.05.030
doi: 10.1016/J.JCMG.2019.05.030
pubmed: 31326477
Xia Y, Chen X, Ravikumar N, et al (2022) Automatic 3D+t four-chamber CMR quantification of the UK biobank: integrating imaging and non-imaging data priors at scale. Med Image Anal 80. https://doi.org/10.1016/J.MEDIA.2022.102498
Bloomgarden DC, Fayad ZA, Ferrari VA et al (1997) Global cardiac function using fast breath-hold MRI: Validation of new acquisition and analysis techniques. Magn Reson Med 37:683–692. https://doi.org/10.1002/MRM.1910370510
doi: 10.1002/MRM.1910370510
pubmed: 9126942
Childs H, Ma L, Ma M et al (2011) Comparison of long and short axis quantification of left ventricular volume parameters by cardiovascular magnetic resonance, with ex-vivo validation. J Cardiovasc Magn Reson 13:1–9. https://doi.org/10.1186/1532-429X-13-40/COMMENTS
doi: 10.1186/1532-429X-13-40/COMMENTS
Fratz S, Chung T, Greil GF et al (2013) Guidelines and protocols for cardiovascular magnetic resonance in children and adults with congenital heart disease: SCMR expert consensus group on congenital heart disease. J Cardiovasc Magn Reson 15:51. https://doi.org/10.1186/1532-429X-15-51
doi: 10.1186/1532-429X-15-51
pubmed: 23763839
pmcid: 3686659