Application of artificial intelligence as postoperative support for patients undergoing thoracic surgery

Artificial intelligence (AI) is playing an increasingly significant role in medicine, from imaging-based diagnosis to personalized treatment. In the postoperative period, it may enhance patient support by answering questions more clearly and reducing the need for in-person visits. However, its reliability and effects on the patient experience remain insufficiently studied. This study examined the potential of AI as a complementary tool in the care of patients undergoing thoracic surgery, assessing response accuracy, patient acceptance, and its potential to reshape postoperative follow-up.

Real patient questions were collected and answered by both ChatGPT and specialist physicians. The responses were compared in terms of clarity, accuracy, and completeness. Participants evaluated their satisfaction with the information provided, their need to seek additional details, and their comfort level with using AI without medical supervision.

Among the patients evaluated, 74.3% reported that ChatGPT fully addressed their questions, and 91.4% found the language clear and accessible. However, 62.8% still indicated a need for medical confirmation. In addition, 51.4% of participants stated that they would seek additional information even after receiving an AI-generated response.

The findings suggest that AI has strong potential to enhance patients' postoperative experience by providing rapid, accessible answers. However, its use should be integrated into a hybrid care model that combines technology with individualized medical oversight.