The role of artificial intelligence in the minimally invasive thoracic surgery: narrative review of the last 15 years

Abstract

Background and Objective: Artificial intelligence (AI) is quickly changing every aspect of everyday life and its application in the medical field is increasing day by day. AI does not yet have a standardized role in clinical practice of many fields, therefore, in this narrative review, we analyzed specifically the state-of-the-art regarding AI application in thoracic surgery, with particular focus on its role in minimally invasive thoracic surgery. Methods: Literature research was performed using the PubMed database and the selection process was performed following Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines, excluding non-English papers, not related to the application of the AI in thoracic surgery and published between 1 January 2010 and 1 April 2025. A total of 603 articles were identified, but at least only 38 were selected as appropriate in terms of specificity and topic. Key Content and Findings: One core principle is the ability of AI to process and interpret vast amounts of data, often exceeding human cognitive capabilities. AI and its sub-sets play a significant role in interpretation of radiological images and three-dimensional (3D) reconstruction, tumor localization and evaluation of thoracic anatomy during surgery. AI algorithms help in identifying high-risk patients and their need of ICU admission, reducing complications and optimizing outcomes. Furthermore, the association between AI and robotic surgery represents a transformative frontier, enhancing intra-operative guidance and decision-making and enabling various levels of autonomy, from intelligent instrument control to semi-autonomous task execution. AI is also providing a high-quality and objective training of young surgeons, evaluating technical skills across various procedures and settings. Conclusions: AI models are significantly improving thoracic surgery by enhancing preoperative assessment, surgical training and RATS but there are still significant hurdles to clear, in particular ethical concerns such as data privacy, who is liable for AI mistakes, algorithmic bias and how it might affect a surgeon’s autonomy.

Ähnliche Arbeiten