RECAP: Towards Precise Radiology Report Generation via Dynamic Disease Progression Reasoning

TL;DR

RECAP is a novel method that enhances radiology report generation by incorporating dynamic disease progression reasoning, precise attribute modeling, and temporal information from historical records to produce more accurate and detailed reports.

Contribution

This work introduces RECAP, a new framework that integrates disease progression graphs and dynamic reasoning to improve the accuracy and precision of radiology report generation.

Findings

Outperforms existing models on two public datasets.

Effectively models disease attributes and progression.

Improves report accuracy and detail.

Abstract

Automating radiology report generation can significantly alleviate radiologists' workloads. Previous research has primarily focused on realizing highly concise observations while neglecting the precise attributes that determine the severity of diseases (e.g., small pleural effusion). Since incorrect attributes will lead to imprecise radiology reports, strengthening the generation process with precise attribute modeling becomes necessary. Additionally, the temporal information contained in the historical records, which is crucial in evaluating a patient's current condition (e.g., heart size is unchanged), has also been largely disregarded. To address these issues, we propose RECAP, which generates precise and accurate radiology reports via dynamic disease progression reasoning. Specifically, RECAP first predicts the observations and progressions (i.e., spatiotemporal information) given two consecutive radiographs. It then combines the historical records, spatiotemporal information, and radiographs for report generation, where a disease progression graph and dynamic progression reasoning mechanism are devised to accurately select the attributes of each observation and progression. Extensive experiments on two publicly available datasets demonstrate the effectiveness of our model.