Transformer-Based Abstract Generation of Medical Case Reports

Abstract

A medical case report gives medical researchers and healthcare providers a thorough account of the symptoms, treatment, and diagnosis of a specific patient. This clinical data is essential because they aid in diagnosing novel or uncommon illnesses, analyzing specific medical occurrences, and enhancing knowledge of current medical education. The summary of the medical case report is needed so that one can decide on further reading as going through the entire contents of a medical case report istime-consuming. In this paper, we present a deep learning methodology for the generation of the automatic summaries of the medical case reports. The final proposed fine-tuned summarizer on the test data set generated a mean precision of 0.4481 and Rouge-1 Score of 0.2803.