Determining temporal information in clinical text
Natural language processing techniques have been successfully developed to extract clinical concepts such as symptoms (hallucinations), treatments (clozapine) and diagnoses (schizophrenia) from health record text such as data from CRIS, but these methods don’t currently capture the timing of these events. We will seek to develop methods that can ascertain when a symptom or treatment happened, finding time information (dates, durations), relevant concepts (symptoms, treatments), and linking these together.