Fig 1

A proposed system in which physician notes are used as input to a model. The output of the model is an imaging protocol, with color coding to denote the significance of terms in influencing the model's decision: red signifies words that negatively influence the prediction, blue denotes the most important words that positively influence the prediction, and white indicates a neutral influence. This system aims to provide a more efficient and accurate method for determining appropriate imaging protocols, while also offering insight into the decision-making process of the model. By incorporating an explainability component, the proposed system has the potential to enhance trust and understanding in the use of machine learning for medical image protocol assignment