Bullet trajectory predicts the need for damage control: an artificial neural network model

Area of Science:

• Trauma Surgery
• Medical Artificial Intelligence
• Ballistics in Medicine

Background:

• Early decision-making for damage control is crucial in trauma care.
• Bullet trajectory has not been previously utilized as a predictor for damage control decisions.
• This study explores the potential of artificial neural networks (ANNs) in predicting damage control needs.

Purpose of the Study:

• To develop and validate an artificial neural network (ANN) model to predict the need for damage control surgery.
• To assess the efficacy of using bullet trajectory and patient blood pressure as input variables for the ANN model.
• To determine if bullet trajectory is a significant factor in the decision to implement damage control.

Main Methods:

• A multilayer perceptron artificial neural network (ANN) was developed using data from 312 patients with abdominal gunshot injuries.
• Input variables included bullet path, trajectory patterns, and admission systolic blood pressure.
• The model's performance was evaluated on prospectively collected data from 34 additional patients.

Main Results:

• The ANN model achieved a high correct classification rate (0.96) and area under the receiver operating characteristic curve (0.94).
• External validation demonstrated a sensitivity of 88% and specificity of 96%.
• Systolic pressure, bullet path crossing the midline, and right upper quadrant trajectory were identified as key predictive factors.

Conclusions:

• Bullet trajectory is a significant and previously unrecognized factor in the decision-making process for damage control in trauma.
• The developed ANN model shows promise for improving the accuracy and timeliness of damage control decisions.
• Incorporating bullet trajectory analysis into trauma protocols could enhance patient outcomes.