Artificial intelligence in thoracic surgery: Perspectives and challenges

Abstract

With the advances of artificial intelligence (AI) in the medical field, particularly the widespread utilization of large language models (LLMs) such as ChatGPT, Claude, Gemini, Llama, and Deepseek, clinical practice is undergoing an unprecedented technological revolution. These cutting-edge technologies facilitate efficient processing and analysis of vast datasets, providing medical professionals with auxiliary diagnoses and treatment suggestions, while markedly enhancing the quality and efficiency of medical services. Over the past decade, the field of thoracic surgery has achieved transformative progress, primarily driven by AI innovations. Consequently, thoracic surgeons must possess a foundational understanding of AI in order to grasp its implications on their daily practice and explore potential ways of integrating this technology into their work. This article reviews the fundamental elements of AI and the relationships between AI-based techniques. It further summarizes the application of AI in thoracic surgery, aiming to enhance thoracic surgeons' comprehensive understanding of the latest developments in this area. Additionally, this article explores the challenges and limitations faced by AI, including data security and privacy concerns, issues of bias and discrimination, challenges in verification and interpretability, ethical and legal considerations, technical obstacles, as well as training and educational requirements. Finally, it explores emerging AI architectures and their paradigm-shifting impacts on medical ecosystems.

Ähnliche Arbeiten