Extracting Thyroid Nodules Characteristics from Ultrasound Reports Using Transformer-based Natural Language Processing Methods

Aman Pathak,Zehao Yu,Daniel Paredes,Elio Paul Monsour,Andrea Ortiz Rocha,Juan P. Brito,Naykky Singh Ospina,Yonghui Wu

2023-04-01

Abstract:The ultrasound characteristics of thyroid nodules guide the evaluation of thyroid cancer in patients with thyroid nodules. However, the characteristics of thyroid nodules are often documented in clinical narratives such as ultrasound reports. Previous studies have examined natural language processing (NLP) methods in extracting a limited number of characteristics (<9) using rule-based NLP systems. In this study, a multidisciplinary team of NLP experts and thyroid specialists, identified thyroid nodule characteristics that are important for clinical care, composed annotation guidelines, developed a corpus, and compared 5 state-of-the-art transformer-based NLP methods, including BERT, RoBERTa, LongFormer, DeBERTa, and GatorTron, for extraction of thyroid nodule characteristics from ultrasound reports. Our GatorTron model, a transformer-based large language model trained using over 90 billion words of text, achieved the best strict and lenient F1-score of 0.8851 and 0.9495 for the extraction of a total number of 16 thyroid nodule characteristics, and 0.9321 for linking characteristics to nodules, outperforming other clinical transformer models. To the best of our knowledge, this is the first study to systematically categorize and apply transformer-based NLP models to extract a large number of clinical relevant thyroid nodule characteristics from ultrasound reports. This study lays ground for assessing the documentation quality of thyroid ultrasound reports and examining outcomes of patients with thyroid nodules using electronic health records.

Computation and Language

What problem does this paper attempt to address?

The problem this paper attempts to address is how to automatically extract clinically relevant feature information from ultrasound reports of thyroid nodules. Specifically, the authors focus on using Transformer-based natural language processing (NLP) methods to identify and extract 16 important features of thyroid nodules, which are crucial for assessing the risk of thyroid cancer. Currently, such information is typically recorded in clinical narrative documents, such as ultrasound reports, and manually extracting this information is both time-consuming and labor-intensive. Therefore, developing an efficient and accurate automated method to extract these features is of great significance for improving the diagnostic quality for patients with thyroid nodules and reducing overdiagnosis.

The paper mentions that although previous studies have explored the use of NLP techniques to extract features of thyroid nodules from ultrasound reports, these studies mainly focused on a few features and mostly employed rule-based systems. In contrast, this study not only covers a broader range of feature categories but also adopts the latest Transformer model, specifically the GatorTron model, which is a large language model trained on a vast amount of clinical text data and has demonstrated better performance. Through this approach, the researchers hope to lay the foundation for subsequent large-scale electronic health record database analyses and patient outcome studies, thereby better understanding the clinical significance of thyroid nodules and providing support for improving patient treatment plans.