Using Natural Language Processing to Identify Low Back Pain in Imaging Reports
A natural language processing (NLP) pipeline was developed to identify lumbar spine imaging findings associated with low back pain (LBP) in X-radiation (X-ray), computed tomography (CT), and magnetic resonance imaging (MRI) reports. A total of 18,640 report datasets were randomly sampled (stratified...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-12-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/12/24/12521 |
| _version_ | 1797461682619940864 |
|---|---|
| author | Yeji Kim Chanyoung Song Gyuseon Song Sol Bi Kim Hyun-Wook Han Inbo Han |
| author_facet | Yeji Kim Chanyoung Song Gyuseon Song Sol Bi Kim Hyun-Wook Han Inbo Han |
| author_sort | Yeji Kim |
| collection | DOAJ |
| description | A natural language processing (NLP) pipeline was developed to identify lumbar spine imaging findings associated with low back pain (LBP) in X-radiation (X-ray), computed tomography (CT), and magnetic resonance imaging (MRI) reports. A total of 18,640 report datasets were randomly sampled (stratified by imaging modality) to obtain a balanced sample of 300 X-ray, 300 CT, and 300 MRI reports. A total of 23 radiologic findings potentially related to LBP were defined, and their presence was extracted from radiologic reports. In developing NLP pipelines, section and sentence segmentation from the radiology reports was performed using a rule-based method, including regular expression with negation detection. Datasets were randomly split into 80% for development and 20% for testing to evaluate the model’s extraction performance. The performance of the NLP pipeline was evaluated by using recall, precision, accuracy, and the F1 score. In evaluating NLP model performances, four parameters—recall, precision, accuracy, and F1 score—were greater than 0.9 for all 23 radiologic findings. These four scores were 1.0 for 10 radiologic findings (listhesis, annular fissure, disc bulge, disc extrusion, disc protrusion, endplate edema or Type 1 Modic change, lateral recess stenosis, Schmorl’s node, osteophyte, and any stenosis). In the seven potentially clinically important radiologic findings, the F1 score ranged from 0.9882 to 1.0. In this study, a rule-based NLP system identifying 23 findings related to LBP from X-ray, CT, and MRI reports was developed, and it presented good performance in regards to the four scoring parameters. |
| first_indexed | 2024-03-09T17:22:55Z |
| format | Article |
| id | doaj.art-4c33fa3811644ae4b099fb587422cd1c |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-09T17:22:55Z |
| publishDate | 2022-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-4c33fa3811644ae4b099fb587422cd1c2023-11-24T13:00:01ZengMDPI AGApplied Sciences2076-34172022-12-0112241252110.3390/app122412521Using Natural Language Processing to Identify Low Back Pain in Imaging ReportsYeji Kim0Chanyoung Song1Gyuseon Song2Sol Bi Kim3Hyun-Wook Han4Inbo Han5Research Competency Milestones Program of School of Medicine, CHA University School of Medicine, Bundang-gu, Seongnam-si 13488, Republic of KoreaDepartment of Biomedical Informatics, CHA University School of Medicine, Bundang-gu, Seongnam-si 13488, Republic of KoreaDepartment of Biomedical Informatics, CHA University School of Medicine, Bundang-gu, Seongnam-si 13488, Republic of KoreaDepartment of Neurosurgery, CHA University School of Medicine, CHA Bungdang Medical Center, Seongnam-si 13497, Republic of KoreaDepartment of Biomedical Informatics, CHA University School of Medicine, Bundang-gu, Seongnam-si 13488, Republic of KoreaDepartment of Neurosurgery, CHA University School of Medicine, CHA Bungdang Medical Center, Seongnam-si 13497, Republic of KoreaA natural language processing (NLP) pipeline was developed to identify lumbar spine imaging findings associated with low back pain (LBP) in X-radiation (X-ray), computed tomography (CT), and magnetic resonance imaging (MRI) reports. A total of 18,640 report datasets were randomly sampled (stratified by imaging modality) to obtain a balanced sample of 300 X-ray, 300 CT, and 300 MRI reports. A total of 23 radiologic findings potentially related to LBP were defined, and their presence was extracted from radiologic reports. In developing NLP pipelines, section and sentence segmentation from the radiology reports was performed using a rule-based method, including regular expression with negation detection. Datasets were randomly split into 80% for development and 20% for testing to evaluate the model’s extraction performance. The performance of the NLP pipeline was evaluated by using recall, precision, accuracy, and the F1 score. In evaluating NLP model performances, four parameters—recall, precision, accuracy, and F1 score—were greater than 0.9 for all 23 radiologic findings. These four scores were 1.0 for 10 radiologic findings (listhesis, annular fissure, disc bulge, disc extrusion, disc protrusion, endplate edema or Type 1 Modic change, lateral recess stenosis, Schmorl’s node, osteophyte, and any stenosis). In the seven potentially clinically important radiologic findings, the F1 score ranged from 0.9882 to 1.0. In this study, a rule-based NLP system identifying 23 findings related to LBP from X-ray, CT, and MRI reports was developed, and it presented good performance in regards to the four scoring parameters.https://www.mdpi.com/2076-3417/12/24/12521natural language processinglow back painlumbar spine imaging |
| spellingShingle | Yeji Kim Chanyoung Song Gyuseon Song Sol Bi Kim Hyun-Wook Han Inbo Han Using Natural Language Processing to Identify Low Back Pain in Imaging Reports Applied Sciences natural language processing low back pain lumbar spine imaging |
| title | Using Natural Language Processing to Identify Low Back Pain in Imaging Reports |
| title_full | Using Natural Language Processing to Identify Low Back Pain in Imaging Reports |
| title_fullStr | Using Natural Language Processing to Identify Low Back Pain in Imaging Reports |
| title_full_unstemmed | Using Natural Language Processing to Identify Low Back Pain in Imaging Reports |
| title_short | Using Natural Language Processing to Identify Low Back Pain in Imaging Reports |
| title_sort | using natural language processing to identify low back pain in imaging reports |
| topic | natural language processing low back pain lumbar spine imaging |
| url | https://www.mdpi.com/2076-3417/12/24/12521 |
| work_keys_str_mv | AT yejikim usingnaturallanguageprocessingtoidentifylowbackpaininimagingreports AT chanyoungsong usingnaturallanguageprocessingtoidentifylowbackpaininimagingreports AT gyuseonsong usingnaturallanguageprocessingtoidentifylowbackpaininimagingreports AT solbikim usingnaturallanguageprocessingtoidentifylowbackpaininimagingreports AT hyunwookhan usingnaturallanguageprocessingtoidentifylowbackpaininimagingreports AT inbohan usingnaturallanguageprocessingtoidentifylowbackpaininimagingreports |