The study is based on the rationale that traditional Mallampatti assessment is subjective and prone to inter-observer variability. By using deep learning techniques, particularly Convolutional Neural Networks (CNNs), the study aims to improve objectivity, accuracy, and consistency in airway assessment. A dataset of approximately 800 annotated oral cavity images will be collected from adult patients, pre-processed, and used to train and validate AI models including ResNet-50 and MobileNetV2. Model performance will be evaluated using accuracy, precision, recall, F1-score, and confusion matrices.

Anticipated outcomes include improved prediction of difficult airways, enhanced clinical decision-making, streamlined preoperative workflows, and potential educational benefits for training clinicians. Ethical considerations such as informed consent, data anonymization, and bias mitigation are emphasized. Overall, the study seeks to establish AI as a reliable and clinically relevant tool for airway assessment in anaesthesiology.