Incorporating Personalization Features in a Hospital-Stay Summary Generation System

Abstract

Most of the currently available health resources contain vast amount of information that are created by keeping the "general" population in mind, which in reality, might not be useful for anyone. One approach to providing comprehensible health information to patients is to generate summaries that are personalized to each individual. This paper details the design of a personalized hospital-stay summary generation system that tailors its content to the patient's understanding of medical terminologies and their level of engagement in improving their own health. Our summaries were found to cover around 80% of the health concepts that were considered as important by a doctor or a nurse. An online survey conducted on 150 participants verified that our algorithm's interpretation of the personalization parameters is representative of that of a larger population.

SATS: simplification aware text summarization of scientific documents

“…In Macdonald and Siddharthan ( 2016 ) simplification was used as a preprocessing step prior to summarization of children's stories. Simplification was also used similarly more recently for clinical data summaries generation (Acharya et al, 2019 ).…”

A Quantitative Analysis of Patients’ Narratives of Heart Failure

“…We are engaged in a large, long-term project that aims to improve patient discharge instructions with a personalized and comprehensible summary of their hospital stay that is informed by the perspectives of the three main stake-holders: doctors, nurses, and patients. Over the last few years, we have developed and implemented a framework for summarizing heterogeneous information (textual discharge notes from the doctor, structured information from the nurses) and providing explanations for difficult medical terms (Di Eugenio et al, 2014;Acharya et al, 2016Acharya et al, , 2018Acharya et al, , 2019. Figure 1 shows a part of a summary that is generated by our system.…”