Artificial Intelligence in Thoracic Surgical Oncology: Technical Foundations and Applications in Lung and Esophageal Cancer Surgery
- 1Tufts Medical Center, Boston, MA 02111, USA
- 2Boston Medical Center Brighton, Brighton, MA 02153, USA
- 3Department of Cardiothoracic Surgery, University of Pittsburgh School of Medicine, University of Pittsburgh Medical Center, Pittsburgh, PA 15213, USA
Abstract
Background/Objectives: Artificial intelligence (AI) is increasingly being investigated in lung and esophageal cancer surgery, where treatment decisions require integrating imaging, clinical risk, and operative findings. This narrative review describes the technical foundations of surgical AI and examines its applications in thoracic surgical oncology, with particular emphasis on robotic surgery and the evidence supporting clinical use. Methods: PubMed/MEDLINE was searched for English-language publications addressing AI, machine learning, deep learning, computer vision, and robotic thoracic surgery. Targeted searches and examination of reference lists identified additional studies. Publications were selected for clinical and technical relevance, prioritizing original investigations, external or prospective validation studies, and reporting guidelines. Applications were examined across preoperative assessment, intraoperative assistance, and postoperative care. Results: Preoperative applications include pulmonary nodule characterization, prediction of nodal disease and invasive tumor features, anatomical reconstruction, and assessment of response to neoadjuvant therapy. Intraoperative systems have demonstrated anatomical segmentation, surgical-phase recognition, and augmented reality guidance. Other models address technical performance, postoperative complications, recurrence, and survival. Selected studies report improved predictive performance or earlier recognition of critical structures; however, much of the evidence remains retrospective and institution-specific. External validation is limited, and evidence that AI assistance reduces complications or improves oncologic outcomes remains insufficient. Conclusions: AI offers tools for interpreting information relevant to thoracic cancer surgery, but technical accuracy alone does not establish clinical benefit. Prospective, multicenter evaluations should determine whether AI changes consequential surgical decisions and improves patient outcomes. Reliable performance under difficult operative conditions, integration into clinical workflow, and surgeon oversight remain essential to adoption.
Related
- The application value of artificial intelligence in medical imaging: foundations, progress, and challenges.
- Artificial intelligence in esophageal cancer: a narrative review from a thoracic surgical perspective
- Artificial intelligence-enabled mental health applications for adolescents: a scoping review of safety, effectiveness and clinical implementation
- Artificial Intelligence Applications in Laser-Assisted Refractive Surgery—A Comprehensive Review
- From underdiagnosis to scalable screening: Applications of artificial intelligence for vertebral compression fracture detection – a systematic review and meta-analysis of diagnostic performance
BibTeX
@article{cassidy2026artificial,
title = {{Artificial Intelligence in Thoracic Surgical Oncology: Technical Foundations and Applications in Lung and Esophageal Cancer Surgery}},
author = {Ryan Cassidy and Sara Razzaq and Austin Howell and Miguel Leiva-Juarez and Lana Schumacher},
journal = {Cancers},
year = {2026},
month = oct,
doi = {10.3390/cancers18213438},
url = {https://www.mdpi.com/2072-6694/18/21/3438},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Artificial Intelligence in Thoracic Surgical Oncology: Technical Foundations and Applications in Lung and Esophageal Cancer Surgery
AU - Ryan Cassidy
AU - Sara Razzaq
AU - Austin Howell
AU - Miguel Leiva-Juarez
AU - Lana Schumacher
PY - 2026
DA - 2026/10/24
JO - Cancers
DO - 10.3390/cancers18213438
UR - https://www.mdpi.com/2072-6694/18/21/3438
AB - Background/Objectives: Artificial intelligence (AI) is increasingly being investigated in lung and esophageal cancer surgery, where treatment decisions require integrating imaging, clinical risk, and operative findings. This narrative review describes the technical foundations of surgical AI and examines its applications in thoracic surgical oncology, with particular emphasis on robotic surgery and the evidence supporting clinical use. Methods: PubMed/MEDLINE was searched for English-language publications addressing AI, machine learning, deep learning, computer vision, and robotic thoracic surgery. Targeted searches and examination of reference lists identified additional studies. Publications were selected for clinical and technical relevance, prioritizing original investigations, external or prospective validation studies, and reporting guidelines. Applications were examined across preoperative assessment, intraoperative assistance, and postoperative care. Results: Preoperative applications include pulmonary nodule characterization, prediction of nodal disease and invasive tumor features, anatomical reconstruction, and assessment of response to neoadjuvant therapy. Intraoperative systems have demonstrated anatomical segmentation, surgical-phase recognition, and augmented reality guidance. Other models address technical performance, postoperative complications, recurrence, and survival. Selected studies report improved predictive performance or earlier recognition of critical structures; however, much of the evidence remains retrospective and institution-specific. External validation is limited, and evidence that AI assistance reduces complications or improves oncologic outcomes remains insufficient. Conclusions: AI offers tools for interpreting information relevant to thoracic cancer surgery, but technical accuracy alone does not establish clinical benefit. Prospective, multicenter evaluations should determine whether AI changes consequential surgical decisions and improves patient outcomes. Reliable performance under difficult operative conditions, integration into clinical workflow, and surgeon oversight remain essential to adoption.
SN - 2072-6694
ER -