Zusammenfassung
Klinisch verfügbare KI(künstliche Intelligenz)-basierte Lösungen dienen der Prozessoptimierung und der Objektivierung bildbasierter Quantifizierung für die kardiale Diagnostik. Damit eröffnet sich eine breitere Anwendung der heutigen Hochleistungsmedizin auch außerhalb von universitären Zentren. In dieser Übersichtsarbeit wird über die Praxisrelevanz sowie die Anwendung der KI in der Bildvorverarbeitung und -segmentierung, der Diagnostik, des Phenotypings für die klinische Entscheidungsunterstützung sowie über den heutigen Stand des Datenschutzes und der interdisziplinären Zusammenarbeit berichtet. Die Weiterentwicklung hochqualitativer KI-basierter Lösungen unter Berücksichtigung aktueller Datenschutzvorgaben erfordert eine multidisziplinäre Zusammenarbeit und den Aufbau geeigneter Forschungsinfrastrukturen. KI-Anwendungen in der diagnostischen kardialen Bildanalyse ohne klinische Daten der Patienten und die letztendlich entscheidende kardiologische Expertise erscheinen für eine valide Befundung und damit Therapieentscheidung bei kardiologischen Patienten nicht sinnvoll.
Abstract
Clinically available artificial intelligence (AI)-based solutions serve to optimize processes and objectify image-based quantification for cardiac diagnostics. This opens up a broader application of today’s high-performance medicine also outside university centers. This overview article reports on the practical relevance and the application of AI in image preprocessing and segmentation, the diagnostics, phenotyping, clinical decision support and on the current status of data protection and interdisciplinary cooperation. The further development of high-quality AI-based solutions, taking current data protection regulations into account, requires multidisciplinary cooperation and the development of suitable research infrastructures. The applications of AI in diagnostic cardiac image analysis without clinical patient data and without the ultimately decisive cardiologic expertise do not appear to make sense for a valid diagnosis and thus treatment decision in cardiologic patients.
Literatur
Bernard O et al (2018) Deep learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: is the problem solved? IEEE Trans Med Imaging 37(11):2514–2525
Friedrich S et al (2021) Applications of artificial intelligence/machine learning approaches in cardiovascular medicine: a systematic review with recommendations. Eur Heart J 2(3):424–436
Ruijsink B et al (2020) Fully automated, quality-controlled cardiac analysis from CMR: validation and large-scale application to characterize cardiac function. JACC Cardiovasc Imaging 13(3):684–695
Zhou J et al (2021) Artificial intelligence in echocardiography: detection, functional evaluation, and disease diagnosis. Cardiovasc Ultrasound 19(1):29
Fotaki A et al (2022) Artificial intelligence in cardiac MRI: is clinical adoption forthcoming? Front Cardiovasc Med. https://doi.org/10.3389/fcvm.2021.818765
Raisi-Estabragh Z et al (2021) Cardiovascular magnetic resonance imaging in the UK Biobank: a major international health research resource. Eur Heart J Cardiovasc Imaging 22(3):251–258
Diaz O et al (2021) Data preparation for artificial intelligence in medical imaging: A comprehensive guide to open-access platforms and tools. Phys Med 83:25–37
Ghorbani A et al (2020) Deep learning interpretation of echocardiograms. NP Digit Med 3:10
Iyer K et al (2021) AngioNet: a convolutional neural network for vessel segmentation in X‑ray angiography. Sci Rep 11(1):18066
Berhane H et al (2020) Fully automated 3D aortic segmentation of 4D flow MRI for hemodynamic analysis using deep learning. Magn Reson Med 84(4):2204–2218
Haji-Valizadeh H et al (2021) Highly accelerated free-breathing real-time phase contrast cardiovascular MRI via complex-difference deep learning. Magn Reson Med 86(2):804–819
Zwanenburg A et al (2020) The image biomarker standardization initiative: standardized quantitative Radiomics for high-throughput image-based phenotyping. Radiology 295(2):328–338
Baessler B et al (2018) Cardiac MRI texture analysis of T1 and T2 maps in patients with infarctlike acute myocarditis. Radiology 289(2):357–365
Mannil M et al (2020) Artificial intelligence and texture analysis in cardiac imaging. Curr Cardiol Rep 22(11):131
Murgia A et al (2020) Cardiac computed tomography radiomics: an emerging tool for the non-invasive assessment of coronary atherosclerosis. Cardiovasc Diagn Ther 10(6):2005–2017
Xu P et al (2021) Radiomics: the next frontier of cardiac computed tomography. Circ Cardiovasc Imaging 14(3):e11747
Huellebrand MI, Tautz L, Kelle S, Hennemuth A (2022) A collaborative approach for the development and application of machine learning solutions for CMR-based cardiac disease classification. Front Cardiovasc Med. https://doi.org/10.3389/fcvm.2022.829512
Tromp J et al (2022) Automated interpretation of systolic and diastolic function on the echocardiogram: a multicohort study. Lancet Digit Health 4(1):e46–e54
Hirata Y et al (2020) Deep learning for screening of pulmonary hypertension using standard chest X‑ray. Eur Heart J 41(Supplement_2):ehaa946.2246
Jeong H‑G et al (2021) Classification of cardioembolic stroke based on a deep neural network using chest radiographs. EBioMedicine 69:103466
Ponikowski P et al (2016) 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail 18(8):891–975
Kameshima H et al (2021) A phenotyping of diastolic function by machine learning improves prediction of clinical outcomes in heart failure. Front Cardiovasc Med 8:755109
Woolley RJ et al (2021) Machine learning based on biomarker profiles identifies distinct subgroups of heart failure with preserved ejection fraction. Eur J Heart Fail 23(6):983–991
Sanchez-Martinez S et al (2018) Machine learning analysis of left ventricular function to characterize heart failure with preserved ejection fraction. Circ Cardiovasc Imaging 11(4):e7138
Assadi H et al (2021) Cardiac MRI for the prognostication of heart failure with preserved ejection fraction: A systematic review and meta-analysis. Magn Reson Imaging 76:116–122
Doeblin P et al (2019) CMR tissue characterization in patients with HFmrEF. J Clin Med 8(11):1877. https://doi.org/10.3390/jcm8111877
Bonde A et al (2021) Assessing the utility of deep neural networks in predicting postoperative surgical complications: a retrospective study. Lancet Digit Health 3(8):e471–e485
Meyer A et al (2018) Machine learning for real-time prediction of complications in critical care: a retrospective study. Lancet Respir Med 6(12):905–914
Nagueh SF et al (2016) Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European association of cardiovascular imaging. J Am Soc Echocardiogr 29(4):277–314
DuMont Schütte A et al (2021) Overcoming barriers to data sharing with medical image generation: a comprehensive evaluation. NPJ Digit Med 4(1):141
Diller G‑P et al (2020) Utility of deep learning networks for the generation of artificial cardiac magnetic resonance images in congenital heart disease. BMC Med Imaging 20(1):113
Ferdian E et al (2022) WSSNet: aortic wall shear stress estimation using deep learning on 4D flow MRI. Front Cardiovasc Med. https://doi.org/10.3389/fcvm.2021.769927
Thamsen B et al (2021) Synthetic database of aortic morphometry and hemodynamics: overcoming medical imaging data availability. IEEE Trans Med Imaging 40(5):1438–1449
Scherer J et al (2020) Joint imaging platform for federated clinical data analytics. JCO Clin Cancer Inform 4:1027–1038
Finanzierung
Diese Arbeit wurde teilweise von der Deutschen Forschungsgemeinschaft (DFG) finanziert – SFB-1470 – B06.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Interessenkonflikt
A. Hennemuth, M. Hüllebrand, P. Doeblin, N. Krüger und S. Kelle geben an, dass kein Interessenkonflikt besteht.
Für diesen Beitrag wurden von den Autoren 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
Hennemuth, A., Hüllebrand, M., Doeblin, P. et al. Anwendungen von künstlicher Intelligenz in der diagnostischen kardialen Bildanalyse. Kardiologe 16, 72–81 (2022). https://doi.org/10.1007/s12181-022-00548-2
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12181-022-00548-2
Schlüsselwörter
- Prozessoptimierung
- Objektivierung bildbasierter Quantifizierung
- Praxisrelevanz
- Datenschutz
- Entscheidungsunterstützung