Multi-Dimensional Evaluation Metrics for Chest X-Ray Reports
In the past few years, there has been abundant research in using machine learning to generate high quality radiology reports using the large MIMIC-CXR chest x-ray dataset. However, there has been little work focused on evaluating the quality of generated reports from a clinical perspective, where ac...
Main Author: | |
---|---|
Other Authors: | |
Format: | Thesis |
Published: |
Massachusetts Institute of Technology
2022
|
Online Access: | https://hdl.handle.net/1721.1/144486 |
_version_ | 1811068707655385088 |
---|---|
author | Rawat, Saumya |
author2 | Szolovits, Peter |
author_facet | Szolovits, Peter Rawat, Saumya |
author_sort | Rawat, Saumya |
collection | MIT |
description | In the past few years, there has been abundant research in using machine learning to generate high quality radiology reports using the large MIMIC-CXR chest x-ray dataset. However, there has been little work focused on evaluating the quality of generated reports from a clinical perspective, where accuracy is the most important factor. Current evaluation metrics evaluate reports in one dimension. This work proposes the use of multiple dimensions (factual correctness, comprehensiveness, style, and overall quality) to better capture evaluation preferences of a clinical text generating model where preferences can differ based on the use case. This work also presents a dataset of radiologist rating annotations for generated and reference chest x-ray radiology reports. Lastly, it also creates an improved metric for the readability dimension by adding context awareness of frequent and acceptable medical terminology. |
first_indexed | 2024-09-23T07:59:55Z |
format | Thesis |
id | mit-1721.1/144486 |
institution | Massachusetts Institute of Technology |
last_indexed | 2024-09-23T07:59:55Z |
publishDate | 2022 |
publisher | Massachusetts Institute of Technology |
record_format | dspace |
spelling | mit-1721.1/1444862022-08-30T04:08:54Z Multi-Dimensional Evaluation Metrics for Chest X-Ray Reports Rawat, Saumya Szolovits, Peter Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science In the past few years, there has been abundant research in using machine learning to generate high quality radiology reports using the large MIMIC-CXR chest x-ray dataset. However, there has been little work focused on evaluating the quality of generated reports from a clinical perspective, where accuracy is the most important factor. Current evaluation metrics evaluate reports in one dimension. This work proposes the use of multiple dimensions (factual correctness, comprehensiveness, style, and overall quality) to better capture evaluation preferences of a clinical text generating model where preferences can differ based on the use case. This work also presents a dataset of radiologist rating annotations for generated and reference chest x-ray radiology reports. Lastly, it also creates an improved metric for the readability dimension by adding context awareness of frequent and acceptable medical terminology. M.Eng. 2022-08-29T15:50:46Z 2022-08-29T15:50:46Z 2022-05 2022-05-27T16:19:36.124Z Thesis https://hdl.handle.net/1721.1/144486 In Copyright - Educational Use Permitted Copyright MIT http://rightsstatements.org/page/InC-EDU/1.0/ application/pdf Massachusetts Institute of Technology |
spellingShingle | Rawat, Saumya Multi-Dimensional Evaluation Metrics for Chest X-Ray Reports |
title | Multi-Dimensional Evaluation Metrics for Chest X-Ray Reports |
title_full | Multi-Dimensional Evaluation Metrics for Chest X-Ray Reports |
title_fullStr | Multi-Dimensional Evaluation Metrics for Chest X-Ray Reports |
title_full_unstemmed | Multi-Dimensional Evaluation Metrics for Chest X-Ray Reports |
title_short | Multi-Dimensional Evaluation Metrics for Chest X-Ray Reports |
title_sort | multi dimensional evaluation metrics for chest x ray reports |
url | https://hdl.handle.net/1721.1/144486 |
work_keys_str_mv | AT rawatsaumya multidimensionalevaluationmetricsforchestxrayreports |