Review Article Hepatic Steatosis Identification: A Comprehensive Approach Using Natural Language Processing and Machine Learning in Radiology Reports
by Oluwafemi O. Balogun1*, Eugenia Uche-Anya1,2,3, Megan E. Michta1, Lisa Mun3, Mariam K. Ibrahim1,2, Emad S. Salman2, Kathleen E. Corey1,2,3
1Liver Center, Division of Gastroenterology, Department of Medicine, Massachusetts General Hospital, Boston Massachusetts, USA
2Harvard Medical School, Boston Massachusetts; Department of Medicine, Massachusetts General Hospital, Boston Massachusetts, USA
3Clinical and Translational Epidemiology Unit (CTEU), Massachusetts General Hospital, Boston Massachusetts, USA
Corresponding author: Balogun OO, Liver Center, Division of Gastroenterology, Department of Medicine, Massachusetts General Hospital, Boston Massachusetts, USA
Received Date: 26 November 2024
Accepted Date: 10 December 2024
Published Date: 13 December 2024.
Citation: Balogun OO, Anya EU, Michta ME, Mun L, Ibrahim MK, et al. (2024) Hepatic Steatosis Identification: A Comprehensive Approach Using Natural Language Processing and Machine Learning in Radiology Reports. J Dig Dis Hepatol 9: 218. https://doi.org/10.29011/2574-3511.100218
Abstract
Manual review of clinical charts and unstructured radiology reports for the finding of fatty infiltration of the liver can be time-consuming, labor-intensive, and inefficient for large retrospective datasets. Abdominal ultrasounds are the most common method of diagnosing fatty liver. In this study, we present a practical alternative approach using Natural Language Processing (NLP) and machine learning to identify hepatic steatosis in abdominal ultrasound scan reports. We employed a Named Entity Recognition (NER) approach, complemented by text featurization techniques and machine learning models. Our method achieved promising results, offering an accurate and efficient way to identify hepatic steatosis in radiology reports.
Keywords: Steatosis; Radiology; Reporting; Natural Language Processing; Artificial Intelligence; MASLD
© by the Authors & Gavin Publishers. This is an Open Access Journal Article Published Under Attribution-Share Alike CC BY-SA: Creative Commons Attribution-Share Alike 4.0 International License. Read More About Open Access Policy.