Specialized Artificial Intelligence Clinical Decision Support System for Aesthetic Surgery: Aesthetic Surgery Using...

Area of Science:

• Medical Informatics
• Plastic Surgery
• Artificial Intelligence in Medicine

Background:

• Traditional literature searches are time-consuming for specific aesthetic surgery clinical questions.
• No AI clinical decision support systems are tailored for evidence-based practice in aesthetic surgery.

Purpose of the Study:

• Develop and evaluate AURA (Aesthetic surgery Using Retrieval Augmentation), a novel AI-powered clinical decision support system.
• Provide evidence-based guidance specifically for aesthetic surgery.

Main Methods:

• Integrated a database of 6,546 aesthetic surgery publications (2001-2024) with a large language model.
• Utilized retrieval-augmented generation technology for information retrieval.
• Assessed system performance across 14 complex clinical scenarios using metrics like faithfulness, relevancy, and correctness.

Main Results:

• AURA achieved high performance scores: 0.94 for faithfulness, 0.86 for relevancy, and 0.77 for correctness.
• Semantic evaluation showed strong results (0.73 SEM Score, 0.80 SEM Max Similarity).
• Performance was better for established techniques than for emerging procedures.

Conclusions:

• AURA is the first specialized AI clinical decision support system for aesthetic surgery.
• The system effectively delivers relevant, accurate, literature-based information.
• AURA offers an efficient, evidence-based resource for aesthetic surgeons' clinical decisions.