A study on comparative evaluation of machine learning algorithms for predicting difficult airways
试验速览
- 阶段
- 不适用
- 状态
- 尚未招募
- 发起方
- 入组人数
- 697
- 试验地点
- 1
- 主要终点
- 1. To compare the predictive performance of various machine learning algorithms in identifying difficult airway cases.
研究概览
简要总结
This study aims to compare different machine learning algorithms in difficult airway prediction (including parameters like BMI , neck circumference, thyromental distance, interincisor gap, mallampatti classification, age and head and neck movements). Using a Macintosh blade of size 3 or 4 , laryngoscopy will be done and vocal cord is graded according to cormack lehane grading. Grades 1 and 2 are considered as easy and 3 and 4 as difficult airways. Five algorithms are systematically compared: Random Forest (RF), Gradient Boosting (GB), XGBoost , Deep Learning (DL) neural network and a Stacking Ensemble combining base models’ predictions.
研究设计
- 研究类型
- Observational
入排标准
- 年龄范围
- 18.00 Year(s) 至 80.00 Year(s)（—）
- 性别
- All
入选标准
- •Surgical procedures requiring endotracheal intubation using the Macintosh blade
- •Age : 18-80years
- •ASA I,II, III.
排除标准
- •Developmental anomalies which may affect airway assessment
- •Patients with airway malformations, midline neck swelling, face trauma or other gross external head and neck deformities
- •Psychiatric patients or patients who are unable to follow commands.
结局指标
主要结局
1. To compare the predictive performance of various machine learning algorithms in identifying difficult airway cases.
时间窗: 3months
次要结局
- 2. To determine the most clinically useful machine learning model for potential integration into preoperative airway assessment workflows.(3 months)
- 3. To identify the optimal subset or combination of predictors that yields the highest predictive accuracy for each machine learning algorithm.(3 months)
研究者
Shahana Muneer
Amala institute of medical sciences