Shimo Lab at "Discharge Me!": Discharge Summarization by Prompt-Driven Concatenation of Electronic Health Record Sections

TL;DR

This paper's approach to the shared task “Discharge Me!” is presented, which involves a first step of extracting the relevant sections from the EHR and then adding explanatory prompts to these sections and concatenate them with separate tokens to create the input text.

Abstract

In this paper, we present our approach to the shared task "Discharge Me!" at the BioNLP Workshop 2024. The primary goal of this task is to reduce the time and effort clinicians spend on writing detailed notes in the electronic health record (EHR). Participants develop a pipeline to generate the "Brief Hospital Course" and "Discharge Instructions" sections from the EHR. Our approach involves a first step of extracting the relevant sections from the EHR. We then add explanatory prompts to these sections and concatenate them with separate tokens to create the input text. To train a text generation model, we perform LoRA fine-tuning on the ClinicalT5-large model. On the final test data, our approach achieved a ROUGE-1 score of $0.394$, which is comparable to the top solutions.

Figures (6)

• Figure 1: Overview of our pipeline. To create input text, we extract sections from the EHR, add explanatory prompts, and then concatenate them with <sep> tokens. We then generate discharge summaries using ClnicalT5-large, which has been fine-tuned for each target.
• Figure 2: An example of the EHR with the location of the target discharge summaries. To show the sections used for the input text, the rounded rectangle is for the "Brief Hospital Course", the dashed rounded rectangle is for the "Discharge Instructions", and the rectangles are for both targets. The symbol "[...]" indicates omissions.
• Figure 3: Brief Hospital Course
• Figure 4: Discharge Instructions
• Figure 6: Brief Hospital Course