Zusammenfassung
Klinisches/methodisches Problem
Kardiale Erkrankungen sind weltweit die führende Todesursache. Viele Erkrankungen können gezielt behandelt werden, sobald eine valide Diagnose gestellt wurde. Die kardiale Magnetresonanztomographie (MRT) hat in der Abklärung zahlreicher kardialer Pathologien einen hohen Stellenwert. Jedoch sind sowohl die Bildakquisition als auch die Befundung und damit zusammenhängende sekundäre Bildauswertungen zeitaufwändig und komplex.
Radiologische Standardverfahren
In den internationalen Leitlinien etabliert sich die kardiale MRT zunehmend in der Evaluation der Herzfunktion und der Differenzialdiagnostik verschiedenster kardialer Erkrankungen.
Methodische Innovationen
Die kardiale MRT besitzt aufgrund der Aufnahmetechnik und der Befundung mit aufwändigen Sekundärmessungen eine eingeschränkte Reproduzierbarkeit. Techniken künstlicher Intelligenz (KI) und Radiomics bieten das Potenzial, die Akquisition, Befundung und Reproduzierbarkeit der kardialen MRT zu verbessern.
Leistungsfähigkeit
Studien zeigen, dass KI und Radiomics-Analysen die kardiale MRT hinsichtlich Bildakquisition, diagnostischer und prognostischer Wertigkeit verbessern können. Zudem konnten mit dieser Herangehensweise neue Biomarker identifiziert werden.
Bewertung und Empfehlung für die Praxis
In der Anwendung von KI in der kardialen MRT liegt großes Potenzial. Die aktuelle Datenlage ist in einigen Aspekten noch zu gering, vor allem liegen zu wenige prospektive und große multizentrische Studien vor. Dadurch sind die entwickelten Algorithmen häufig wissenschaftlich nicht ausreichend validiert und finden in der klinischen Routine noch keine Anwendung.
Abstract
Clinical/methodical issue
Cardiac diseases are the leading cause of death. Many diseases can be specifically treated once a valid diagnosis is established. Cardiac magnetic resonance imaging (MRI) plays a central role in the workup of many cardiac pathologies. However, image acquisition as well as interpretation and related secondary image evaluation are time-consuming and complex.
Standard radiological methods
Cardiac MRI is becoming increasingly established in international guidelines for the evaluation of cardiac function and differential diagnosis of a wide variety of cardiac diseases.
Methodological innovations
Cardiac MRI has limited reproducibility due to the acquisition technique and interpretation of findings with complex secondary measurements. Artificial intelligence techniques and radiomics offer the potential to improve the acquisition, interpretation, and reproducibility of cardiac MRI.
Performance
Research suggests that artificial intelligence and radiomic analysis can improve cardiac MRI in terms of image acquisition and also diagnostic and prognostic value. Furthermore, the implementation of artificial intelligence and radiomics may result in the identification of new biomarkers.
Achievements and practical recommendations
The implementation of artificial intelligence in cardiac MRI has great potential. However, the current level of evidence is still limited in some aspects; in particular there are too few prospective and large multicenter studies available. As a result, the algorithms developed are often not sufficiently validated scientifically and are not yet applied in clinical routine.
Literatur
Knuuti J, Wijns W, Saraste A et al (2020) 2019 ESC Guidelines for the diagnosis and management of chronic coronary syndromes. Eur Heart J 41:407–477. https://doi.org/10.1093/eurheartj/ehz425
McDonagh TA, Metra M, Adamo M et al (2021) 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J 42:3599–3726. https://doi.org/10.1093/eurheartj/ehab368
Petersen SE, Aung N, Sanghvi MM et al (2017) Reference ranges for cardiac structure and function using cardiovascular magnetic resonance (CMR) in Caucasians from the UK Biobank population cohort. J Cardiovasc Magn Reson 19:18. https://doi.org/10.1186/s12968-017-0327-9
Choy G, Khalilzadeh O, Michalski M et al (2018) Current applications and future impact of machine learning in radiology. Radiology 288:318–328. https://doi.org/10.1148/radiol.2018171820
Aerts HJWL, Velazquez ER, Leijenaar RTH et al (2014) Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun 5:4006. https://doi.org/10.1038/ncomms5006
Hauptmann A, Arridge S, Lucka F et al (2019) Real-time cardiovascular MR with spatio-temporal artifact suppression using deep learning-proof of concept in congenital heart disease. Magn Reson Med 81:1143–1156. https://doi.org/10.1002/mrm.27480
Blansit K, Retson T, Masutani E et al (2019) Deep learning-based prescription of cardiac MRI planes. Radiol Artif Intell 1:e180069. https://doi.org/10.1148/ryai.2019180069
Schuppert C, von Krüchten R, Hirsch JG et al (2022) Whole-body magnetic resonance imaging in the large population-based German national cohort study: predictive capability of automated image quality assessment for protocol repetitions. Invest Radiol 57:478–487. https://doi.org/10.1097/RLI.0000000000000861
Zhang N, Yang G, Gao Z et al (2019) Deep learning for diagnosis of chronic myocardial infarction on nonenhanced cardiac cine MRI. Radiology 291:606–617. https://doi.org/10.1148/radiol.2019182304
Xia Y, Ravikumar N, Greenwood JP et al (2021) Super-resolution of cardiac MR cine imaging using conditional GANs and unsupervised transfer learning. Med Image Anal 71:102037. https://doi.org/10.1016/j.media.2021.102037
Bernard O, Lalande A, Zotti C et al (2018) Deep learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: Is the problem solved? IEEE Trans Med Imaging 37:2514–2525. https://doi.org/10.1109/TMI.2018.2837502
Bai W, Sinclair M, Tarroni G et al (2018) Automated cardiovascular magnetic resonance image analysis with fully convolutional networks. J Cardiovasc Magn Reson 20:65. https://doi.org/10.1186/s12968-018-0471-x
Davies RH, Augusto JB, Bhuva A et al (2022) Precision measurement of cardiac structure and function in cardiovascular magnetic resonance using machine learning. J Cardiovasc Magn Reson 24:16. https://doi.org/10.1186/s12968-022-00846-4
Zheng Q, Delingette H, Duchateau N, Ayache N (2018) 3‑D consistent and robust segmentation of cardiac images by deep learning with spatial propagation. IEEE Trans Med Imaging 37:2137–2148. https://doi.org/10.1109/TMI.2018.2820742
Chen D, Bhopalwala H, Dewaswala N et al (2022) Deep neural network for cardiac magnetic resonance image segmentation. J Imaging. https://doi.org/10.3390/jimaging8050149
Peng P, Lekadir K, Gooya A et al (2016) A review of heart chamber segmentation for structural and functional analysis using cardiac magnetic resonance imaging. Magma 29:155–195. https://doi.org/10.1007/s10334-015-0521-4
Winther HB, Hundt C, Schmidt B et al (2018) ν‑net: deep learning for generalized biventricular mass and function parameters using multicenter cardiac MRI data. JACC Cardiovasc Imaging 11:1036–1038. https://doi.org/10.1016/j.jcmg.2017.11.013
Fahmy AS, Rausch J, Neisius U et al (2018) Automated cardiac MR scar quantification in hypertrophic cardiomyopathy using deep convolutional neural networks. JACC Cardiovasc Imaging 11:1917–1918. https://doi.org/10.1016/j.jcmg.2018.04.030
Fahmy AS, Rowin EJ, Chan RH et al (2021) Improved quantification of myocardium scar in late gadolinium enhancement images: deep learning based image fusion approach. J Magn Reson Imaging 54:303–312. https://doi.org/10.1002/jmri.27555
Neisius U, El-Rewaidy H, Nakamori S et al (2019) Radiomic analysis of myocardial native T1 imaging discriminates between hypertensive heart disease and hypertrophic cardiomyopathy. JACC Cardiovasc Imaging 12:1946–1954. https://doi.org/10.1016/j.jcmg.2018.11.024
Schofield R, Ganeshan B, Kozor R et al (2016) CMR myocardial texture analysis tracks different etiologies of left ventricular hypertrophy. J Cardiovasc Magn Reson 18:O82. https://doi.org/10.1186/1532-429X-18-S1-O82
Baessler B, Luecke C, Lurz J et al (2019) Cardiac MRI and texture analysis of myocardial T1 and T2 maps in myocarditis with acute versus chronic symptoms of heart failure. Radiology 292:608–617. https://doi.org/10.1148/radiol.2019190101
Chen B‑H, An D‑A, He J et al (2021) Myocardial extracellular volume fraction radiomics analysis for differentiation of reversible versus irreversible myocardial damage and prediction of left ventricular adverse remodeling after ST-elevation myocardial infarction. Eur Radiol 31:504–514. https://doi.org/10.1007/s00330-020-07117-9
Dawes TJW, de Marvao A, Shi W et al (2017) Machine learning of three-dimensional right ventricular motion enables outcome prediction in pulmonary hypertension: a cardiac MR imaging study. Radiology 283:381–390. https://doi.org/10.1148/radiol.2016161315
Kwak S, Everett RJ, Treibel TA et al (2021) Markers of myocardial damage predict mortality in patients with aortic stenosis. J Am Coll Cardiol 78:545–558. https://doi.org/10.1016/j.jacc.2021.05.047
Krebs J, Mansi T, Delingette H et al (2021) CinE caRdiac magneTic resonAnce to predIct veNTricular arrhYthmia (CERTAINTY). Sci Rep 11:22683. https://doi.org/10.1038/s41598-021-02111-7
Raisi-Estabragh Z, Jaggi A, Gkontra P et al (2021) Cardiac magnetic resonance radiomics reveal differential impact of sex, age, and vascular risk factors on cardiac structure and myocardial tissue. Front Cardiovasc Med 8:763361. https://doi.org/10.3389/fcvm.2021.763361
Ponsiglione A, Stanzione A, Cuocolo R et al (2022) Cardiac CT and MRI radiomics: systematic review of the literature and radiomics quality score assessment. Eur Radiol 32:2629–2638. https://doi.org/10.1007/s00330-021-08375-x
Bamberg F, Kauczor H‑U, Weckbach S et al (2015) Whole-body MR imaging in the German national cohort: rationale, design, and technical background. Radiology 277:206–220. https://doi.org/10.1148/radiol.2015142272
Puyol-Anton E, Ruijsink B, Gerber B et al (2019) Regional multi-view learning for cardiac motion analysis: application to identification of dilated cardiomyopathy patients. IEEE Trans Biomed Eng 66:956–966. https://doi.org/10.1109/TBME.2018.2865669
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Interessenkonflikt
A. Rau, M. Soschynski, J. Taron, P. Ruile, C.L. Schlett, F. Bamberg und T. Krauss geben an, dass kein Interessenkonflikt besteht.
Für diesen Beitrag wurden von den Autor/-innen keine Studien an Menschen oder Tieren durchgeführt. Für die aufgeführten Studien gelten die jeweils dort angegebenen ethischen Richtlinien.
Additional information
QR-Code scannen & Beitrag online lesen
Rights and permissions
About this article
Cite this article
Rau, A., Soschynski, M., Taron, J. et al. Künstliche Intelligenz und Radiomics. Radiologie 62, 947–953 (2022). https://doi.org/10.1007/s00117-022-01060-0
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00117-022-01060-0