Mechanical ventilation (MV) is one of the most important life-support systems that's widely used in clinics and hospitals, with the goal of predicting the correct pressure applied to intubed patient's lungs. Despite it's popularity the techonology constituting MV has been unchange for several years,thus it's does not contain any intelligence method to calculate the correct pressure to be applied.
In this project, we make use of Deep Neural Networks to create a model that predicts the correct pressure to be applied on patient's lungs during a breath cycle (inhalation - exhalation). The model takes into account structual factors of the patient's lungs in order to make a prediction.
You can read the paper about this project here.
To train and test the model we used artificial data from kaggle's Google Brain competition which can be found here.
To train your own model from scratch clone this repository and run train.py which can be found in src folder.