Inspecting state of the art performance and NLP metrics in image-based medical report generation
TL
TL;DR
It is shown that simple and even naive approaches yield near SOTA performance on most traditional NLP metrics, and that evaluation methods in this task should be further studied towards correctly measuring clinical accuracy, involving physicians to contribute to this end.
Abstract
Several deep learning architectures have been proposed over the last years to deal with the task of generating a written report given an imaging exam as input. Most works evaluate the generated reports using standard Natural Language Processing (NLP) metrics (e.g. BLEU, ROUGE), reporting significant progress. This article contrast this progress by comparing state of the art (SOTA) models against weak baselines. We show that simple and even naive approaches yield near SOTA performance on most traditional NLP metrics. We conclude that evaluation methods in this task should be further studied towards correctly measuring clinical accuracy, involving physicians to contribute to this end.
References18 items
1
Tangled up in BLEU: Reevaluating the Evaluation of Automatic Machine Translation Evaluation Metrics
2
Baselines for Chest X-Ray Report Generation
3
When Radiology Report Generation Meets Knowledge Graph
4
Optimizing the Factual Correctness of a Summary: A Study of Summarizing Radiology Reports
5