1
|
Koo BS, Jang M, Oh JS, Shin K, Lee S, Joo KB, Kim N, Kim TH. Machine learning models with time-series clinical features to predict radiographic progression in patients with ankylosing spondylitis. JOURNAL OF RHEUMATIC DISEASES 2024; 31:97-107. [PMID: 38559800 PMCID: PMC10973352 DOI: 10.4078/jrd.2023.0056] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/04/2023] [Revised: 10/15/2023] [Accepted: 10/30/2023] [Indexed: 04/04/2024]
Abstract
Objective Ankylosing spondylitis (AS) is chronic inflammatory arthritis causing structural damage and radiographic progression to the spine due to repeated and continuous inflammation over a long period. This study establishes the application of machine learning models to predict radiographic progression in AS patients using time-series data from electronic medical records (EMRs). Methods EMR data, including baseline characteristics, laboratory findings, drug administration, and modified Stoke AS Spine Score (mSASSS), were collected from 1,123 AS patients between January 2001 and December 2018 at a single center at the time of first (T1), second (T2), and third (T3) visits. The radiographic progression of the (n+1)th visit (Pn+1=(mSASSSn+1-mSASSSn)/(Tn+1-Tn)≥1 unit per year) was predicted using follow-up visit datasets from T1 to Tn. We used three machine learning methods (logistic regression with the least absolute shrinkage and selection operation, random forest, and extreme gradient boosting algorithms) with three-fold cross-validation. Results The random forest model using the T1 EMR dataset best predicted the radiographic progression P2 among the machine learning models tested with a mean accuracy and area under the curves of 73.73% and 0.79, respectively. Among the T1 variables, the most important variables for predicting radiographic progression were in the order of total mSASSS, age, and alkaline phosphatase. Conclusion Prognosis predictive models using time-series data showed reasonable performance with clinical features of the first visit dataset when predicting radiographic progression.
Collapse
Affiliation(s)
- Bon San Koo
- Department of Internal Medicine, Inje University Ilsan Paik Hospital, Inje University College of Medicine, Seoul, Korea
| | - Miso Jang
- Department of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
- Department of Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Ji Seon Oh
- Department of Information Medicine, Big Data Research Center, Asan Medical Center, Seoul, Korea
| | - Keewon Shin
- Department of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Seunghun Lee
- Department of Radiology, Hanyang University Hospital for Rheumatic Diseases, Seoul, Korea
| | - Kyung Bin Joo
- Department of Radiology, Hanyang University Hospital for Rheumatic Diseases, Seoul, Korea
| | - Namkug Kim
- Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
- Department of Convergence Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Tae-Hwan Kim
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, Korea
| |
Collapse
|
2
|
Kim TH, Park SY, Shin JH, Lee S, Joo KB, Koo BS. Association between changes in serum alkaline phosphatase levels and radiographic progression in ankylosing spondylitis. Sci Rep 2023; 13:9093. [PMID: 37277451 DOI: 10.1038/s41598-023-36340-9] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/11/2022] [Accepted: 06/01/2023] [Indexed: 06/07/2023] Open
Abstract
This retrospective study evaluated the electronic medical records of patients with ankylosing spondylitis (AS) (January 2001-December 2018) to determine the relationship between serum alkaline phosphatase (ALP) levels and radiographic changes over time. Longitudinal data, including serum ALP levels, were imputed by linear interpolation at 3-month intervals. Among the serum ALP levels calculated for 8 years prior to modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) measurement, those having the highest beta coefficient with the mSASSS were selected in the correlation between ALP and longitudinal mSASSS. Linear mixed models with the selected serum ALP levels, mSASSS, and clinical variables were investigated. We included 1122 patients (mean follow-up, 8.20 [standard deviation: 2.85] years). The serum ALP level from 5 years and 3 months prior showed the highest beta coefficient with the mSASSS. In the linear mixed model, the serum ALP level at 5 years and 3 months before radiographic changes was significantly associated with the mSASSS (β = 0.021, 95% confidence interval: 0.017-0.025, p < 0.001). Serum ALP levels measured approximately 5 years before may be a surrogate marker for predicting spinal radiographic changes. Long-term prospective clinical and experimental studies of > 5 years are required for biomarker discovery or therapeutic research on AS radiographic progression.
Collapse
Affiliation(s)
- Tae-Hwan Kim
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| | - Seo Young Park
- Department of Statistics and Data Science, Korea National Open University, Seoul, South Korea
| | - Ji Hui Shin
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| | - Seunghun Lee
- Department of Radiology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| | - Kyung Bin Joo
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| | - Bon San Koo
- Division of Rheumatology, Department of Internal Medicine, Inje University Seoul Paik Hospital, Inje University College of Medicine, 9, Mareunnae-ro, Jung-gu, Seoul, 04551, South Korea.
| |
Collapse
|
3
|
Kang J, Lee TH, Park SY, Lee S, Koo BS, Kim TH. Prediction of radiographic progression pattern in patients with ankylosing spondylitis using group-based trajectory modeling and decision trees. Front Med (Lausanne) 2022; 9:994308. [PMID: 36341272 PMCID: PMC9631932 DOI: 10.3389/fmed.2022.994308] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/14/2022] [Accepted: 10/07/2022] [Indexed: 11/12/2023] Open
Abstract
OBJECTIVE This study aimed to identify trajectories of radiographic progression of the spine over time and use them, along with associated clinical factors, to develop a prediction model for patients with ankylosing spondylitis (AS). METHODS Data from the medical records of patients diagnosed with AS in a single center were extracted between 2001 and 2018. Modified Stoke Ankylosing Spondylitis Spinal Scores (mSASSS) were estimated from cervical and lumbar radiographs. Group-based trajectory modeling classified patients into trajectory subgroups using longitudinal mSASSS data. In multivariate analysis, significant clinical factors associated with trajectories were selected and used to develop a decision tree for prediction of radiographic progression. The most appropriate group for each patient was then predicted using decision tree analysis. RESULTS We identified three trajectory classes: class 1 had a uniformly increasing slope of mSASSS, class 2 showed sustained low mSASSS, and class 3 showed little change in the slope of mSASSS but highest mSASSS from time of diagnosis to after progression. In multivariate analysis for predictive factors, female sex, younger age at diagnosis, lack of eye involvement, presence of peripheral joint involvement, and low baseline erythrocyte sedimentation rate (log) were significantly associated with class 2. Class 3 was significantly associated with male sex, older age at diagnosis, presence of ocular involvement, and lack of peripheral joint involvement when compared with class 1. Six clinical factors from multivariate analysis were used for the decision tree for classifying patients into three trajectories of radiographic progression. CONCLUSION We identified three patterns of radiographic progression over time and developed a decision tree based on clinical factors to classify patients according to their trajectories of radiographic progression. Clinically, this model holds promise for predicting prognosis in patients with AS.
Collapse
Affiliation(s)
- Juyeon Kang
- Division of Rheumatology, Department of Internal Medicine, Inje University Busan Paik Hospital, Inje University College of Medicine, Busan, South Korea
| | - Tae-Han Lee
- Department of Internal Medicine, Kyungpook National University Chilgok Hospital, Daegu, South Korea
| | - Seo Young Park
- Department of Statistics and Data Science, Korea National Open University, Seoul, South Korea
| | - Seunghun Lee
- Department of Radiology, Hanyang University College of Medicine, Hanyang University Seoul Hospital, Seoul, South Korea
| | - Bon San Koo
- Division of Rheumatology, Department of Internal Medicine, Inje University Seoul Paik Hospital, Inje University College of Medicine, Seoul, South Korea
| | - Tae-Hwan Kim
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| |
Collapse
|
4
|
Kwon SR, Kim TH, Kim TJ, Park W, Shim SC. The Epidemiology and Treatment of Ankylosing Spondylitis in Korea. JOURNAL OF RHEUMATIC DISEASES 2022; 29:193-199. [PMID: 37476425 PMCID: PMC10351411 DOI: 10.4078/jrd.22.0023] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/28/2022] [Revised: 09/16/2022] [Accepted: 09/19/2022] [Indexed: 07/22/2023]
Abstract
Ankylosing spondylitis is a chronic inflammatory disorder characterized by inflammation of the axial skeleton and sacroiliac joints and to a lesser extent by peripheral arthritis and the involvement of some extra-articular organs. It is paramount for the provision of effective health care delivery to be familiar with the epidemiologic studies on prevalence, mortality, and disability. Furthermore, there is no systematic arrangement of studies related to the treatment of ankylosing spondylitis in Korea. In this review, we addressed Korean ankylosing spondylitis epidemiological studies related to prevalence, genetic factor especially human leucocyte antigen-B27, extra-articular manifestations, infections, mortality, radiologic progression, child-birth, and quality of life. Furthermore, we reviewed Korean ankylosing spondylitis treatment researches about treatment trend, patients' registration program called The KOrean College of Rheumatology BIOlogics and targeted therapy (KOBIO) registry project, biologics and biosimiliars, complications especially infections, and issues about bony progression. There would be value to further studying the epidemiology and treatment of Korean ankylosing spondylitis.
Collapse
Affiliation(s)
- Seong-Ryul Kwon
- Rheumatism Center, Department of Internal Medicine, Inha University College of Medicine, Incheon, Korea
| | - Tae-Hwan Kim
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, Korea
| | - Tae-Jong Kim
- Department of Rheumatology, Chonnam National University Medical School and Hospital, Gwangju, Korea
| | - Won Park
- Rheumatism Center, Department of Internal Medicine, Inha University College of Medicine, Incheon, Korea
| | - Seung Cheol Shim
- Division of Rheumatology, Regional Rheumatoid and Degenerative Arthritis Center, Chungnam National University Hospital, Daejeon, Korea
| |
Collapse
|
5
|
Koo BS, Lee JJ, Jung JW, Kang CH, Joo KB, Kim TH, Lee S. A pilot study on deep learning-based grading of corners of vertebral bodies for assessment of radiographic progression in patients with ankylosing spondylitis. Ther Adv Musculoskelet Dis 2022; 14:1759720X221114097. [PMID: 35898565 PMCID: PMC9310199 DOI: 10.1177/1759720x221114097] [Citation(s) in RCA: 7] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/17/2022] [Accepted: 06/29/2022] [Indexed: 11/25/2022] Open
Abstract
Background: Radiographs are widely used to evaluate radiographic progression with
modified stoke ankylosing spondylitis spinal score (mSASSS). Objective: This pilot study aimed to develop a deep learning model for grading the
corners of the cervical and lumbar vertebral bodies for computer-aided
detection of mSASSS in patients with ankylosing spondylitis (AS). Methods: Digital radiographic examination of the spine was performed using Discovery
XR656 (GE Healthcare) and Digital Diagnost (Philips). The disk points were
detected between the bodies using a key-point detection deep learning model
from the image obtained in DICOM (digital imaging and communications in
medicine) format from the cervical and lumbar spinal radiographs. After
cropping the vertebral regions around the disk point, the lower and upper
corners of the vertebral bodies were classified as grade 3 (total bony
bridges) or grades 0, 1, or 2 (non-bridges). We trained a convolutional
neural network model to predict the grades in the lower and upper corners of
the vertebral bodies. The performance of the model was evaluated in a
validation set, which was separate from the training set. Results: Among 1280 patients with AS for whom mSASSS data were available, 5,083
cervical and 5245 lumbar lateral radiographs were reviewed. The total number
of corners where mSASSS was measured in the cervical and lumbar vertebrae,
including the upper and lower corners, was 119,414. Among them, the number
of corners in the training and validation sets was 110,088 and 9326,
respectively. The mean accuracy, sensitivity, and specificity for mSASSS
scoring in one corner of the vertebral body were 0.91604, 0.80288, and
0.94244, respectively. Conclusion: A high-performance deep learning model for grading the corners of the
vertebral bodies was developed for the first time. This model must be
improved and further validated to develop a computer-aided tool for
assessing mSASSS in the future.
Collapse
Affiliation(s)
- Bon San Koo
- Division of Rheumatology, Department of Internal Medicine, Inje University Seoul Paik Hospital, College of Medicine, Inje University, Seoul, Korea
| | | | | | - Chang Ho Kang
- Department of Radiology, Korea University Anam Hospital, Seoul, Korea
| | - Kyung Bin Joo
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, Korea
| | - Tae-Hwan Kim
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, Korea
| | - Seunghun Lee
- Department of Radiology, Hanyang University Hospital for Rheumatic Diseases, 222-1, Wangsimni-ro, Seongdong-gu, Seoul 04763, Korea
| |
Collapse
|
6
|
Lee TH, Koo BS, Nam B, Kim YJ, Son D, Lee S, Joo KB, Kim TH. Age-stratified trends in the progression of spinal radiographic damage in patients with ankylosing spondylitis: a longitudinal study. Ther Adv Musculoskelet Dis 2022; 14:1759720X221100301. [PMID: 35634353 PMCID: PMC9131377 DOI: 10.1177/1759720x221100301] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/01/2022] [Accepted: 04/26/2022] [Indexed: 12/17/2022] Open
Abstract
Objective: The objective of this study was to investigate spinal radiographic progression in specific age ranges of ankylosing spondylitis (AS) patients. Methods: Longitudinal data for 1125 AS patients at a single hospital from 2000 to 2018 were retrospectively reviewed. Radiographic intervals were obtained from patients with consecutive spinal radiographs. The radiographic progression rate was defined as the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) change per year within each interval. Using generalized estimating equations (GEEs), estimated marginal means were calculated for the mSASSS progression rate across age groups after adjusting for potential confounders. Results: We obtained 4016 radiographic intervals and stratified them into five groups based on patient age at the interval start: <20 (n = 122); 20–29 (n = 1124); 30–39 (n = 1690); 40–49 (n = 794); and ⩾50 years (n = 286). The mean (SD) mSASSS progression rate for all the intervals was 0.8 (1.9). The GEE-estimated mean mSASSS progression rate increased with age, peaking in the 30–39 age group with a value of 1.15 [95% confidence interval (CI) 1.03, 1.27], and decreased slightly thereafter. In the presence of risk factors, rapid progression occurred at earlier ages: the GEE-estimated mean mSASSS progression rate in those with elevated C-reactive protein levels and preexisting syndesmophytes was 2.82 (95% CI 1.93, 3.71) in the 20–29 age group. Conclusion: Spinal structural damage in AS seems to progress most rapidly when patients are age 30–39 years. An awareness of the trends in radiographic progression with advancing age could improve understanding of the natural course of AS.
Collapse
Affiliation(s)
- Tae-Han Lee
- Department of Rheumatology, Keimyung University Dongsan Hospital, Daegu, South Korea
| | - Bon San Koo
- Department of Internal Medicine, Inje University Seoul Paik Hospital, Seoul, South Korea
| | - Bora Nam
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| | - Yun Jin Kim
- Biostatistical Consulting and Research Lab, Medical Research Collaborating Center, Hanyang University, Seoul, South Korea
| | - Donghee Son
- Biostatistical Consulting and Research Lab, Medical Research Collaborating Center, Hanyang University, Seoul, South Korea
| | - Seunghun Lee
- Department of Radiology, Hanyang University Hospital, Seoul, South Korea
| | - Kyung Bin Joo
- Department of Radiology, Hanyang University Hospital, Seoul, South Korea
| | - Tae-Hwan Kim
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, 222-1, Wangsimni-ro, Seongdong-gu, Seoul 04763, South Korea
| |
Collapse
|
7
|
2022 French Society for Rheumatology (SFR) recommendations on the everyday management of patients with spondyloarthritis, including psoriatic arthritis. Joint Bone Spine 2022; 89:105344. [PMID: 35038574 DOI: 10.1016/j.jbspin.2022.105344] [Citation(s) in RCA: 36] [Impact Index Per Article: 18.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 11/24/2021] [Indexed: 11/22/2022]
Abstract
OBJECTIVE Update the French Society for Rheumatology (SFR) recommendations on the everyday management of patients with spondyloarthritis, including psoriatic arthritis. METHODS Following standardized procedures, a systematic literature review was done by four supervised rheumatology residents based on questions defined by a task force of 16 attending rheumatologists. The findings were reviewed during three working meetings that culminated in each recommendation receiving a grade and the level of agreement among experts being determined. RESULTS Five general principles and 15 recommendations were developed. They take into account pharmacological and non-pharmacological measures along with treatment methods based on the dominant phenotype present (axial, articular, enthesitis/dactylitis) and the extra-articular manifestations (psoriasis, inflammatory bowel disease, uveitis). NSAIDs are the first-line pharmacological treatment in the various presentations. Conventional synthetic disease modifying antirheumatic drugs (csDMARDs) are not indicated in the axial and isolated entheseal forms. If the response to conventional treatment is not adequate, targeted therapies (biologics, synthetics) should be considered; the indications depend on the clinical phenotype and presence of extra-articular manifestations. CONCLUSION This update incorporates recent data (published since the prior update in 2018) and the predominant clinical phenotype concept. It aims to help physicians with the everyday management of patients affected by spondyloarthritis, including psoriatic arthritis.
Collapse
|
8
|
Koo BS, Lee S, Oh JS, Park SY, Ahn GY, Shin JH, Joo KB, Kim TH. Early control of C-reactive protein levels with non-biologics is associated with slow radiographic progression in radiographic axial spondyloarthritis. Int J Rheum Dis 2021; 25:311-316. [PMID: 34935282 DOI: 10.1111/1756-185x.14268] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/16/2021] [Revised: 11/29/2021] [Accepted: 12/06/2021] [Indexed: 12/22/2022]
Abstract
AIM Predicting radiographic progression is vital for assessing the prognosis of patients with radiographic axial spondyloarthritis, and C-reactive protein (CRP) may be a valuable biomarker for this purpose. This study aimed to investigate the relationship between changes in the CRP level and spinal radiographic progression in patients with radiographic axial spondyloarthritis who were initially treated with non-biologics. METHODS Patients with radiographic axial spondyloarthritis who were followed up for 18 years at a single center and initially treated with nonsteroidal anti-inflammatory drugs and/or conventional disease-modifying antirheumatic drugs for 3 months were included. Patients with a CRP level of <0.8 mg/dL or 50% of the baseline CRP at 3 months were assigned to the controlled CRP group (n = 351), and the remaining patients were assigned to the uncontrolled CRP group (n = 452). A generalized estimating equation was used to analyze the differences in the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) between the 2 groups. RESULTS The increase in the mSASSS was slower in the controlled CRP group than in the uncontrolled CRP group (interaction term β = -.499, 95% confidence interval -0.699 to -0.300). CONCLUSION Controlled CRP achieved in response to initial treatment with non-biologic agents for 3 months was significantly associated with a slower rate of spinal radiographic change in patients with radiographic axial spondyloarthritis. The CRP level at 3 months after initial non-biologic treatment is a good predictor of radiographic progression.
Collapse
Affiliation(s)
- Bon San Koo
- Department of Internal Medicine, Inje University Seoul Paik Hospital, Inje University College of Medicine, Seoul, South Korea
| | - Seunghun Lee
- Department of Radiology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| | - Ji Seon Oh
- Department of Information Medicine, Big Data Research Center, Asan Medical Center, Seoul, South Korea
| | - Seo Young Park
- Department of Statistics and Data Science, Korea National Open University, Seoul, South Korea
| | - Ga Young Ahn
- Division of Rheumatology, Department of Internal Medicine, Korea University Guro Hospital, Seoul, South Korea
| | - Ji Hui Shin
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| | - Kyung Bin Joo
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| | - Tae-Hwan Kim
- Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, South Korea
| |
Collapse
|