1
|
Habibi MA, Naseri Alavi SA, Soltani Farsani A, Mousavi Nasab MM, Tajabadi Z, Kobets AJ. Predicting the Outcome and Survival of Patients with Spinal Cord Injury Using Machine Learning Algorithms: A Systematic Review. World Neurosurg 2024; 188:150-160. [PMID: 38796146 DOI: 10.1016/j.wneu.2024.05.103] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/10/2024] [Accepted: 05/16/2024] [Indexed: 05/28/2024]
Abstract
BACKGROUND Spinal cord injury (SCI) is a significant public health issue, leading to physical, psychological, and social complications. Machine learning (ML) algorithms have shown potential in diagnosing and predicting the functional and neurologic outcomes of subjects with SCI. ML algorithms can predict scores for SCI classification systems and accurately predict outcomes by analyzing large amounts of data. This systematic review aimed to examine the performance of ML algorithms for diagnosing and predicting the outcomes of subjects with SCI. METHODS The literature was comprehensively searched for the pertinent studies from inception to May 25, 2023. Therefore, electronic databases of PubMed, Embase, Scopus, and Web of Science were systematically searched with individual search syntax. RESULTS A total of 9424 individuals diagnosed with SCI across multiple studies were analyzed. Among the 21 studies included, 5 specifically aimed to evaluate diagnostic accuracy, while the remaining 16 focused on exploring prognostic factors or management strategies. CONCLUSIONS ML and deep learning (DL) have shown great potential in various aspects of SCI. ML and DL algorithms have been employed multiple times in predicting and diagnosing patients with SCI. While there are studies on diagnosing acute SCI using DL algorithms, further research is required in this area.
Collapse
Affiliation(s)
- Mohammad Amin Habibi
- Skull Base Research Center, Loghman Hakim Hospital, Shahid Beheshti University of Medical Sciences, Tehran, Iran; Clinical Research Development Center, Faculty of Medicine, Qom University of Medical Sciences, Qom, Iran
| | | | | | | | - Zohreh Tajabadi
- Digestive Disease Research Institute, Tehran University of Medical Sciences, Tehran, Iran
| | - Andrew J Kobets
- Department of Neurological Surgery, Montefiore Medical, Bronx, NY, USA
| |
Collapse
|
2
|
Holtenius J, Mosfeldt M, Enocson A, Berg HE. Prediction of mortality among severely injured trauma patients A comparison between TRISS and machine learning-based predictive models. Injury 2024; 55:111702. [PMID: 38936227 DOI: 10.1016/j.injury.2024.111702] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/13/2024] [Revised: 06/13/2024] [Accepted: 06/19/2024] [Indexed: 06/29/2024]
Abstract
BACKGROUND Given the huge impact of trauma on hospital systems around the world, several attempts have been made to develop predictive models for the outcomes of trauma victims. The most used, and in many studies most accurate predictive model, is the "Trauma Score and Injury Severity Score" (TRISS). Although it has proven to be fairly accurate and is widely used, it has faced criticism for its inability to classify more complex cases. In this study, we aimed to develop machine learning models that better than TRISS could predict mortality among severely injured trauma patients, something that has not been studied using data from a nationwide register before. METHODS Patient data was collected from the national trauma register in Sweden, SweTrau. The studied period was from the 1st of January 2015 to 31st of December 2019. After feature selection and multiple imputation of missing data three machine learning (ML) methods (Random Forest, eXtreme Gradient Boosting, and a Generalized Linear Model) were used to create predictive models. The ML models and TRISS were then tested on predictive ability for 30-day mortality. RESULTS The ML models were well-calibrated and outperformed TRISS in all the tested measurements. Among the ML models, the eXtreme Gradient Boosting model performed best with an AUC of 0.91 (0.88-0.93). CONCLUSION This study showed that all the developed ML-based prediction models were superior to TRISS for the prediction of trauma mortality.
Collapse
Affiliation(s)
- Jonas Holtenius
- Department of Clinical Science, Intervention and Technology, Karolinska Institute, 14152 Stockholm, Sweden; Department of Trauma, Acute Surgery and Orthopaedics, Karolinska University Hospital, 17177 Stockholm, Sweden.
| | - Mathias Mosfeldt
- Department of Molecular Medicine and Surgery, Karolinska Institute, 17176 Stockholm, Sweden; Department of Trauma, Acute Surgery and Orthopaedics, Karolinska University Hospital, 17177 Stockholm, Sweden
| | - Anders Enocson
- Department of Molecular Medicine and Surgery, Karolinska Institute, 17176 Stockholm, Sweden; Department of Trauma, Acute Surgery and Orthopaedics, Karolinska University Hospital, 17177 Stockholm, Sweden
| | - Hans E Berg
- Department of Clinical Science, Intervention and Technology, Karolinska Institute, 14152 Stockholm, Sweden; Department of Trauma, Acute Surgery and Orthopaedics, Karolinska University Hospital, 17177 Stockholm, Sweden
| |
Collapse
|
3
|
Kim Y, Lim M, Kim SY, Kim TU, Lee SJ, Bok SK, Park S, Han Y, Jung HY, Hyun JK. Integrated Machine Learning Approach for the Early Prediction of Pressure Ulcers in Spinal Cord Injury Patients. J Clin Med 2024; 13:990. [PMID: 38398304 PMCID: PMC10889422 DOI: 10.3390/jcm13040990] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/19/2023] [Revised: 01/19/2024] [Accepted: 02/07/2024] [Indexed: 02/25/2024] Open
Abstract
(1) Background: Pressure ulcers (PUs) substantially impact the quality of life of spinal cord injury (SCI) patients and require prompt intervention. This study used machine learning (ML) techniques to develop advanced predictive models for the occurrence of PUs in patients with SCI. (2) Methods: By analyzing the medical records of 539 patients with SCI, we observed a 35% incidence of PUs during hospitalization. Our analysis included 139 variables, including baseline characteristics, neurological status (International Standards for Neurological Classification of Spinal Cord Injury [ISNCSCI]), functional ability (Korean version of the Modified Barthel Index [K-MBI] and Functional Independence Measure [FIM]), and laboratory data. We used a variety of ML methods-a graph neural network (GNN), a deep neural network (DNN), a linear support vector machine (SVM_linear), a support vector machine with radial basis function kernel (SVM_RBF), K-nearest neighbors (KNN), a random forest (RF), and logistic regression (LR)-focusing on an integrative analysis of laboratory, neurological, and functional data. (3) Results: The SVM_linear algorithm using these composite data showed superior predictive ability (area under the receiver operating characteristic curve (AUC) = 0.904, accuracy = 0.944), as demonstrated by a 5-fold cross-validation. The critical discriminators of PU development were identified based on limb functional status and laboratory markers of inflammation. External validation highlighted the challenges of model generalization and provided a direction for future research. (4) Conclusions: Our study highlights the importance of a comprehensive, multidimensional data approach for the effective prediction of PUs in patients with SCI, especially in the acute and subacute phases. The proposed ML models show potential for the early detection and prevention of PUs, thus contributing substantially to improving patient care in clinical settings.
Collapse
Affiliation(s)
- Yuna Kim
- Department of Rehabilitation Medicine, College of Medicine, Dankook University, Cheonan 31116, Republic of Korea; (Y.K.); (S.Y.K.); (T.U.K.); (S.J.L.)
| | - Myungeun Lim
- Digital Biomedical Research Division, Electronics and Telecommunications Research Institute, Daejeon 34129, Republic of Korea; (M.L.); (S.P.); (Y.H.)
| | - Seo Young Kim
- Department of Rehabilitation Medicine, College of Medicine, Dankook University, Cheonan 31116, Republic of Korea; (Y.K.); (S.Y.K.); (T.U.K.); (S.J.L.)
| | - Tae Uk Kim
- Department of Rehabilitation Medicine, College of Medicine, Dankook University, Cheonan 31116, Republic of Korea; (Y.K.); (S.Y.K.); (T.U.K.); (S.J.L.)
| | - Seong Jae Lee
- Department of Rehabilitation Medicine, College of Medicine, Dankook University, Cheonan 31116, Republic of Korea; (Y.K.); (S.Y.K.); (T.U.K.); (S.J.L.)
| | - Soo-Kyung Bok
- Department of Rehabilitation Medicine, College of Medicine, Chungnam National University, Daejeon 35015, Republic of Korea;
| | - Soojun Park
- Digital Biomedical Research Division, Electronics and Telecommunications Research Institute, Daejeon 34129, Republic of Korea; (M.L.); (S.P.); (Y.H.)
| | - Youngwoong Han
- Digital Biomedical Research Division, Electronics and Telecommunications Research Institute, Daejeon 34129, Republic of Korea; (M.L.); (S.P.); (Y.H.)
| | - Ho-Youl Jung
- Digital Biomedical Research Division, Electronics and Telecommunications Research Institute, Daejeon 34129, Republic of Korea; (M.L.); (S.P.); (Y.H.)
| | - Jung Keun Hyun
- Department of Rehabilitation Medicine, College of Medicine, Dankook University, Cheonan 31116, Republic of Korea; (Y.K.); (S.Y.K.); (T.U.K.); (S.J.L.)
- Department of Nanobiomedical Science and BK21 NBM Global Research Center for Regenerative Medicine, Dankook University, Cheonan 31116, Republic of Korea
- Institute of Tissue Regeneration Engineering, Dankook University, Cheonan 31116, Republic of Korea
| |
Collapse
|
4
|
Karabacak M, Jagtiani P, Margetis K. The Predictive Abilities of Machine Learning Algorithms in Patients with Thoracolumbar Spinal Cord Injuries. World Neurosurg 2024; 182:e67-e90. [PMID: 38030070 DOI: 10.1016/j.wneu.2023.11.043] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/01/2023] [Revised: 11/08/2023] [Accepted: 11/09/2023] [Indexed: 12/01/2023]
Abstract
OBJECTIVES The goal of this study is to implement machine learning (ML) algorithms to predict mortality, non-home discharge, prolonged length of stay (LOS), prolonged length of intensive care unit stay (ICU-LOS), and major complications in patients diagnosed with thoracolumbar spinal cord injury, while creating a publicly accessible online tool. METHODS The American College of Surgeons Trauma Quality Program database was used to identify patients with thoracolumbar spinal cord injury. Feature selection was performed with the Least Absolute Shrinkage and Selection Operator algorithm. Five ML algorithms, including TabPFN, TabNet, XGBoost, LightGBM, and Random Forest, were used along with the Optuna optimization library for hyperparameter tuning. RESULTS A total of 147,819 patients were included in the analysis. For each outcome, we determined the best model for deployment in our web application based on the area under the receiver operating characteristic (AUROC) values. The top performing algorithms were as follows: LightGBM for mortality with an AUROC of 0.885, TabPFN for non-home discharge with an AUROC of 0.801, LightGBM for prolonged LOS with an AUROC of 0.673, Random Forest for prolonged ICU-LOS with an AUROC of 0.664, and LightGBM for major complications with an AUROC of 0.73. CONCLUSIONS ML models demonstrate good predictive ability for in-hospital mortality and non-home discharge, fair predictive ability for major complications and prolonged ICU-LOS, but poor predictive ability for prolonged LOS. We have developed a web application that allows these models to be accessed.
Collapse
Affiliation(s)
- Mert Karabacak
- Department of Neurosurgery, Mount Sinai Health System, New York, New York, USA
| | - Pemla Jagtiani
- School of Medicine, SUNY Downstate Health Sciences University, New York, New York, USA
| | | |
Collapse
|
5
|
Wasiak K, Frasuńska J, Tarnacka B. Can the Initial Parameters of Functional Scales Predict Recovery in Patients with Complete Spinal Cord Injury? A Retrospective Cohort Study. Diagnostics (Basel) 2024; 14:129. [PMID: 38248006 PMCID: PMC10814489 DOI: 10.3390/diagnostics14020129] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/10/2023] [Revised: 12/31/2023] [Accepted: 01/03/2024] [Indexed: 01/23/2024] Open
Abstract
Regaining greater independence in performing daily activities constitutes a priority for people with tetraplegia following spinal cord injury (SCI). The highest expectations are connected with the improvement of hand function. Therefore, it is so important for the clinician to identify reliable and commonly applicable prognostic factors for functional improvement. The aim of this study was to conduct an analysis to assess the impact of initial functional factors on the clinical improvement in patients during early neurological rehabilitation (ENR). This study assessed 38 patients with complete SCI aged 17-78 who underwent ENR in 2012-2022. The analysis included the motor score from the AIS (MS), the Barthel Index (BI) and the SCIM scale values at the beginning of the ENR program and after its completion. During ENR, patients achieved a statistically significant improvement in MS, BI and SCIM. The initial MS and the level of neurological injury constituted the predictors of functional improvement during ENR. Significant statistical relationships were observed primarily in the correlations between the initial MS and BI, and the increase in the analyzed functional scales of SCI patients. Higher initial MS may increase the chances of a greater and faster functional improvement during ENR.
Collapse
Affiliation(s)
- Krzysztof Wasiak
- Department of Rehabilitation, Mazovian Rehabilitation Center STOCER, 05-520 Konstancin-Jeziorna, Poland;
| | - Justyna Frasuńska
- Department of Rehabilitation, Medical University of Warsaw, 02-637 Warsaw, Poland;
| | - Beata Tarnacka
- Department of Rehabilitation, Medical University of Warsaw, 02-637 Warsaw, Poland;
- Department of Rehabilitation, National Institute of Geriatrics, Rheumatology and Rehabilitation, 02-637 Warsaw, Poland
| |
Collapse
|
6
|
Zhou W, Liu Y, Wang Z, Mao Z, Li M. Serum glucose/potassium ratio as a clinical risk factor for predicting the severity and prognosis of acute traumatic spinal cord injury. BMC Musculoskelet Disord 2023; 24:870. [PMID: 37946195 PMCID: PMC10633987 DOI: 10.1186/s12891-023-07013-5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 07/21/2023] [Accepted: 11/02/2023] [Indexed: 11/12/2023] Open
Abstract
OBJECTIVE Acute traumatic Spinal cord injury (TSCI) is a devastating event that causes severe sensory and motor impairments as well as autonomic dysfunction in patients, yet relevant clinical biomarkers have not been established. This study aimed to determine the significance of the serum glucose/potassium ratio (GPR) in evaluating TSCI severity and predicting prognosis. METHODS An analysis of 520 clinical records of acute TSCI patients from January 2012 to June 2022 was conducted. The relationships between serum GPR and The American Spinal Injury Association Impairment Scale (AIS) grade 6-month post-trauma prognosis and the admission AIS grade were analyzed. To evaluate the discriminatory ability, a receiver operating characteristic curve (ROC) analysis was used. All methods were performed in accordance with the relevant guidelines and regulations. RESULTS Based on the initial assessment of AIS grade, 256 (49.2%) patients were categorized into the severe TSCI group (AIS A-B), and there was a significant correlation between the severe TSCI group and serum GPR (p < 0.001). Serum GPR was reduced in an AIS grade-dependent manner (R = - 0.540, p < 0.001). Of the 520 patients, 262 (50.4%) patients were classified as having a poor prognosis according to the AIS grade at discharge. Serum GPR was also reduced in an AIS grade at discharge-dependent manner (R = - 0.599, p < 0.001), and was significantly higher in the poor prognosis group compared to the good prognosis group (p < 0.001). Poor prognosis was significantly associated with sex (p = 0.009), severity of TSCI (p < 0.001), location of TSCI (p < 0.001), surgical decompression (p < 0.018), body temperature (p < 0.001), heart rate (p < 0.001), systolic arterial pressure (SAP) (p < 0.001), diastolic arterial pressure (DAP) (p < 0.001), serum GPR (p < 0.001), serum glucose (p < 0.001), serum potassium (p < 0.001), and white blood cell count (p = 0.003). Multivariate logistic regression analysis showed a significant correlation between poor prognosis and serum GPR (p = 0.023). The ROC analysis showed the area under the curve of serum GPR to be a poor predictor of prognosis in TSCI patients at 0.842 (95% confidence interval, 0.808-0.875). CONCLUSION There was a significant relationship between serum GPR and admission injury severity and the 6-month prognosis of acute TSCI patients. Serum GPR serves as a readily available clinical risk factor for predicting the severity and 6-month prognosis of acute traumatic spinal cord injury, which holds potential clinical significance for patients with TSCI.
Collapse
Affiliation(s)
- Wu Zhou
- Department of Neurosurgery, The First Affiliated Hospital of Nanchang University, 17 Yongwai Street, Nanchang, 330006, Jiangxi, China
| | - Yihao Liu
- Department of Neurosurgery, The First Affiliated Hospital of Nanchang University, 17 Yongwai Street, Nanchang, 330006, Jiangxi, China
| | - Zhihua Wang
- Department of Neurosurgery, The First Affiliated Hospital of Nanchang University, 17 Yongwai Street, Nanchang, 330006, Jiangxi, China
| | - Zelu Mao
- Department of Neurosurgery, The First Affiliated Hospital of Nanchang University, 17 Yongwai Street, Nanchang, 330006, Jiangxi, China
| | - Meihua Li
- Department of Neurosurgery, The First Affiliated Hospital of Nanchang University, 17 Yongwai Street, Nanchang, 330006, Jiangxi, China.
| |
Collapse
|