Weaning patients from mechanical ventilation (MV) is a crucial process that requires meticulous assessment of their ability to breathe independently without ventilatory assistance.

Spontaneous breathing test (SBT) is a widely used method for this assessment, involving minimal or no assistance breathing for a defined duration. Satisfactory tolerance of SBT indicates that the patient is ready for extubation.

However, this decision is not straightforward, as it requires balancing the risks of weaning failure with those of complications from prolonged MV.

Therefore, objective tools are essential to reinforce clinical decision making, improving the safety and efficacy of the weaning process.

The article presents a convolutional neural network (CNN)-based classification system designed to determine a patient’s suitability for extubation after SBT.

CNNs are a category of artificial neural networks known for their convolutional layers that employ filters to extract relevant features from the input.

Read More