An AI based algorithm and framework for automatic classification of events in an incident reporting system.
Sun Z, Zong W, Miller B, Chetty I, and Wen N. An ai based algorithm and framework for automatic classification of events in an incident reporting system. Med Phys 2018; 45(6):e377.
Purpose: Clinical incident reporting, tracking, and risk analysis are critical parts of quality assurance (QA) workflow. Traditional incident reporting and learning systems rely heavily on the user input to identify, report, investigate, categorize, and respond to incidents. We have developed and trained an artificial intelligence-guided incident learning application utilizing recursive neural network (RNN) to automatically determine categories and severity level of incidents, and identify risks in clinical workflows. Methods: 2617 existing incident tickets from 19 categories entered by clinical staff were used for development of the classification model. 90% of the events were used for training and 10% for validation of the RNN-based algorithm. A subset of 355 tickets with severity level using a 0-10 scale were labeled by physicists for severity estimator model training and validation. Text messages from the training tickets were preprocessed with tokenization, stop words and punctuation removal, and lemmatisation, and were fed into the RNN. We used bidirectional RNN-based gated recurrent units (GRUs) architecture for wordlevel processing. Outputs of the forward and backward RNNs were then fed into a softmax output layer for multi-label classification and a linear output layer for regression analysis of the severity level. The AI model were deployed as REST API services which served as an engine to analyze incidents entered through a web application. Immediate feedback is provided through a web interface. Results: After 50 epochs of training, the RNN model reached accuracy of 84.4%, 78.0%, 68.3% for classification of the top three, two and one categories, respectively. RMS error of the severity level prediction reached 0.79 after 74 epochs. Conclusion: We developed a RNNbased incident learning framework for automatic multi-label classification and severity level estimation. Accuracy of the AI-based algorithm is likely to improve with a larger number of events, and more refined categorization of events used for training.