ICU embedding (AI in healthcare)

Abstract

Intensive Care Units (ICUs) collect vast amounts of data, including patient vitals, often recorded as time series data. This data can be analyzed to extract meaningful data, which can aid in timely intervention, improving patient outcomes. The complexity and quantity of this data also make it impossible to analyze manually. Traditional statistical methods are not adequate to extract meaningful information and recognize patterns. Deep neural network based language models; however, they excel at identifying patterns. This project leverages this to train a BERT-based [2] transformer for Masked Language Modelling (MLM) and Next Sentence Prediction (NSP) tasks. Procuring the embeddings from this transformer can aid further research on this topic.