Related Experiment Video
Updated: Jan 12, 2026

Drug-Induced Sleep Endoscopy DISE with Target Controlled Infusion TCI and Bispectral Analysis in Obstructive Sleep Apnea
Published on: December 6, 2016
Experts V/S AI´s 2.0: Comparative evaluation of AI models and expert consensus in obstructive sleep apnea assessment
Giovanni Cammaroto1,2, Felipe Ahumada Mira2,3, Valentin Favier2,4
1Head and Neck Department, ENT & Oral Surgery Unity, G.B. Morgagni, L. Pierantoni Hospital, Forlì, Italy.
Artificial intelligence chatbots show promise in managing obstructive sleep apnea (OSA), but expert human judgment remains crucial for complex cases. ChatGPT-3.5 surprisingly outperformed newer AI models in this study.
Area of Science:
- Medical Informatics
- Artificial Intelligence in Medicine
- Sleep Medicine
Background:
- Obstructive sleep apnea (OSA) management requires expert clinical judgment.
- Artificial intelligence (AI) chatbots are increasingly explored for medical applications.
- Evaluating AI performance in specialized medical domains is essential.
Purpose of the Study:
- To compare the diagnostic and management recommendations of AI chatbots for obstructive sleep apnea (OSA) against expert otolaryngologist consensus.
- To assess the performance of ChatGPT-3.5, ChatGPT-4.0, and Gemini in responding to OSA-related clinical scenarios.
- To determine the utility of AI as a complementary tool in OSA management.
Main Methods:
- A 10-question survey on OSA management was administered to ChatGPT-3.5, ChatGPT-4.0, Gemini, and 100 otolaryngologists.
- Responses were evaluated by ten sleep medicine super-experts using a Likert scale for agreement with expert consensus.
- Statistical analyses were conducted to compare AI performance against human expert consensus.
Main Results:
- Expert consensus achieved the highest score (4.5 ± 0.9), significantly outperforming all AI models.
- ChatGPT-3.5 (4.1 ± 1.2) performed best among AI, followed by ChatGPT-4.0 (3.9 ± 1.4) and Gemini (3.6 ± 1.5).
- AI models showed agreement in specific scenarios but struggled with complex cases requiring multi-factor integration, particularly in therapeutic management.
Conclusions:
- AI models show potential as supportive tools for OSA management in well-defined scenarios.
- Human expert clinical judgment remains indispensable for complex OSA cases.
- ChatGPT-3.5's superior performance over newer versions highlights that general AI model updates do not always translate to better specialized medical domain performance.
More Related Videos
Related Concept Videos
Sleep Apnea
The condition is more prevalent among...
Assessment of Airway, Skin Color, and Use of Accessory Muscles
Introduction
The initial evaluation of a patient's respiratory system...
Cardiopulmonary Resuscitation II: ACLS Airway Management
Physical Assessment of the Respiratory Tract II: Inspection
Chest Configuration
The chest configuration...
Assessment of Ventilation I: Respiratory Rate
A Ventilation assessment is critical for monitoring a patient's health status. Respiration, one of the most accessible vital signs, provides insights into the function of numerous body systems and can indicate serious health issues, such as brainstem injuries from head trauma.
Critical Guidelines for Assessing Ventilation:

