# Joint Imbalance Adaptation for Radiology Report Generation

## TL;DR

This paper introduces a new model for radiology report generation that addresses data imbalance issues to improve report accuracy.

## Contribution

The novel Joint Imbalance Adaptation (JIMA) model uses a curriculum learning strategy to handle imbalanced data in radiology reports.

## Key findings

- JIMA improves evaluation metrics by 16.75% to 50.50% on radiology datasets.
- The model's focus on rare labels and clinical tokens enhances clinical accuracy in reports.
- Curriculum learning reduces overfitting to frequent patterns and underfitting to rare ones.

## Abstract

Radiology report generation, translating radiological images into precise and clinically relevant description, may face the data imbalance challenge – medical tokens appear less frequently than regular tokens; and normal entries are significantly more than abnormal ones. However, very few studies consider the imbalance issues, not even with conjugate imbalance factors.

In this study, we propose a Joint Imbalance Adaptation (JIMA) model to promote task robustness by leveraging token and label imbalance. JIMA predicts entity distributions from images and generates reports based on these distributions and image features. We employ a hard-to-easy learning strategy that mitigates overfitting to frequent labels and tokens, thereby encouraging the model to focus more on rare labels and clinical tokens.

JIMA shows notable improvements (16.75% - 50.50% on average) across evaluation metrics on IU X-ray and MIMIC-CXR datasets. Our ablation analysis proves that JIMA’s enhanced handling of infrequent tokens and abnormal labels counts the major contribution. Human evaluation and case study experiments further validate that JIMA can generate more clinically accurate reports.

Data imbalance (e.g., infrequent tokens and abnormal labels) leads to the underperformance of radiology report generation. Our curriculum learning strategy successfully reduce data imbalance impacts by reducing overfitting on frequent patterns and underfitting on infrequent patterns. While data imbalance remains challenging, our approach opens new directions for the generation task.