Artificial Intelligence in Cardiothoracic Surgery: Promises, Failures, and the Gap to Clinical Reality
Medical Sciences · 01.11.2026 · 14 (7) · 734
DOI: 10.3390/medsci14070734 ↗Kurz gefasst
Fachartikel in der Zeitschrift Medical Sciences, online erschienen am 01.11.2026. Verfasst von Vasileios Leivaditis und 8 weiteren Autor:innen. Die Arbeit nennt 67 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background
The growing interest in artificial intelligence (AI) in cardiothoracic surgery has led to a substantial increase in scientific publications, although its integration into routine clinical practice remains limited. Despite encouraging developments in several areas of surgical care, the clinical benefits of many proposed applications have yet to be established.
Objective
This review critically examines the discrepancy between the reported capabilities of AI-based technologies and their implementation in clinical cardiothoracic surgical practice, with particular emphasis on the factors limiting their broader adoption.
Methods
A semi-systematic literature review covering publications from January 2018 through September 2026 was conducted to examine the development, validation, and clinical application of AI-based tools across the preoperative, intraoperative, and postoperative stages of cardiothoracic surgical care. Particular attention was given to methodological limitations, external validation, and barriers to clinical implementation. Key Findings: Although several AI models have demonstrated promising performance in retrospective studies and controlled research settings, evidence supporting their reliability and clinical utility in routine practice remains limited. Important challenges include insufficient external and prospective validation, variable generalizability across clinical settings, limited model interpretability, and difficulties integrating AI-based systems into established surgical and intensive care workflows. Regulatory requirements, technical infrastructure, and human factors represent additional barriers to implementation.
Conclusions
The clinical translation of AI in cardiothoracic surgery requires more than improvements in algorithmic performance. Greater emphasis on robust external validation, prospective clinical evaluation, workflow integration, and appropriate regulatory oversight is necessary to establish the safety, effectiveness, and practical value of these technologies. Addressing these challenges will be essential for translating advances in AI research into meaningful improvements in cardiothoracic surgical care.
Zitieren
APA
Leivaditis, V., Maniatopoulos, A. A., Karangelis, D., Androutsopoulou, V., Panagiotopoulos, I., Papadima, A., Papathanassiou, Z., Nikolakopoulos, K., & Mulita, F. (2026). Artificial Intelligence in Cardiothoracic Surgery: Promises, Failures, and the Gap to Clinical Reality. Medical Sciences, 14(7), 734. https://doi.org/10.3390/medsci14070734
Vancouver
Leivaditis V, Maniatopoulos AA, Karangelis D, Androutsopoulou V, Panagiotopoulos I, Papadima A, et al. Artificial Intelligence in Cardiothoracic Surgery: Promises, Failures, and the Gap to Clinical Reality. Medical Sciences. 2026;14(7):734. doi:10.3390/medsci14070734
BibTeX
@article{Leivaditis2026,
author = {Leivaditis, Vasileios and Maniatopoulos, Andreas Antonios and Karangelis, Dimos and Androutsopoulou, Vasiliki and Panagiotopoulos, Ioannis and Papadima, Artemisia and Papathanassiou, Zafeiria and Nikolakopoulos, Konstantinos and Mulita, Francesk},
title = {Artificial Intelligence in Cardiothoracic Surgery: Promises, Failures, and the Gap to Clinical Reality},
journal = {Medical Sciences},
year = {2026},
volume = {14},
number = {7},
pages = {734},
doi = {10.3390/medsci14070734}
}
EUVIMED