Machine learning in clinical diagnosis, prognostication, and management of acute traumatic spinal cord injury (SCI): A systematic review

Dietz, Nicholas; Jaganathan, V.; Alkin, Victoria; Mettille, Jersey; Boakye, Maxwell; Drazin, Doniel

doi:10.1016/j.jcot.2022.102046

Cited by 23 publications

(15 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Artificial neural networks and machine learning models have assumed paramount significance in numerous applications that profoundly affect human activities. These methodologies have gained prominence due to their exceptional predictive accuracy [1][2][3][4][5][6]. Neuronal networks, now named deep learning, re-emerged after 2010 due to massive improvements in computer resources, some innovations, and successful applications [7].…”

Section: Introductionmentioning

confidence: 99%

Machine Learning Training Optimization using the Barycentric Correction Procedure

Ramos-Pulido,

Hernández-Gress,

G. CeballosCancino

2024

Artificial Intelligence and Big Data

View full text Add to dashboard Cite

Machine learning (ML) algorithms are predictively competitive algorithms with many humanimpact applications. However, the issue of long execution time remains unsolved in the literature for high-dimensional spaces. This study proposes combining ML algorithms with an efficient methodology known as the barycentric correction procedure (BCP) to address this issue. This study uses synthetic data and an educational dataset from a private university to show the benefits of the proposed method. It was found that this combination provides significant benefits related to time in synthetic and real data without losing accuracy when the number of instances and dimensions increases. Additionally, for high-dimensional spaces, it was proved that BCP and linear support vector classification (LinearSVC), after an estimated feature map for the gaussian radial basis function (RBF) kernel, were unfeasible in terms of computational time and accuracy.

show abstract

Section: Introductionmentioning

confidence: 99%

Machine Learning Training Optimization using the Barycentric Correction Procedure

Ramos-Pulido,

Hernández-Gress,

G. CeballosCancino

2024

Artificial Intelligence and Big Data

View full text Add to dashboard Cite

show abstract

“…The surgical implants in postoperative MRI scans might also cause severe image artifacts. Deep learning (DL) can improve the diagnosis and prognostication in SCI by automating the lesion annotation process, thereby reducing rater-specific biases and facilitating the analysis of large SCI cohorts across sites (21)(22)(23). Indeed, quantitative SCI lesion biomarkers derived from DL-based automatic segmentations have been shown to correlate well with clinical measures of motor impairment (24).…”

Section: Introductionmentioning

confidence: 99%

“…Indeed, quantitative SCI lesion biomarkers derived from DL-based automatic segmentations have been shown to correlate well with clinical measures of motor impairment (20). However, despite its numerous advantages, DL has not been sufficiently explored in the context of traumatic SCI (18), with no open-source methods existing to date. This suggests a need for an automatic biomarker identification method that deals with the complex pathophysiology of traumatic SCI patients, generalizes to multiple sites and is easily accessible by researchers.…”

Section: Introductionmentioning

confidence: 99%

“…As a result, most studies involve manual identification of these biomarkers (6,8,(13)(14)(15)(16), which is a time-consuming process susceptible to inter-rater variability, making it less reproducible in multi-site studies. Deep learning (DL) can improve the diagnosis and prognostication in SCI by automating the lesion annotation process, thereby reducing rater-specific biases and facilitating the analysis of large SCI cohorts across sites (17)(18)(19). Indeed, quantitative SCI lesion biomarkers derived from DL-based automatic segmentations have been shown to correlate well with clinical measures of motor impairment (20).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

SCIseg: Automatic Segmentation of T2-weighted Intramedullary Lesions in Spinal Cord Injury

Naga Karthik,

Valosek,

Smith

et al. 2024

Preprint

View full text Add to dashboard Cite

BackgroundQuantitative MRI biomarkers in spinal cord injury (SCI) can help understand the extent of the focal injury. However, due to the lack of automatic segmentation methods, these biomarkers are derived manually, which is a time-consuming process prone to intra- and inter-rater variability, thus limiting large multi-site studies and translation to clinical workflows.PurposeTo develop a deep learning tool for the automatic segmentation of T2-weighted hyperintense lesions and the spinal cord in SCI patients.Material and MethodsThis retrospective study included a cohort of SCI patients from three sites enrolled between July 2002 and February 2023 who underwent clinical MRI examination. A deep learning model,SCIseg, was trained on T2-weighted images with heterogeneous image resolutions (isotropic, anisotropic), and orientations (axial, sagittal) acquired using scanners from different manufacturers (Siemens, Philips, GE) and different field strengths (1T, 1.5T, 3T) for the automatic segmentation of SCI lesions and the spinal cord. The proposed method was visually and quantitatively compared with other open-source baseline methods. Quantitative biomarkers (lesion volume, lesion length, and maximal axial damage ratio) computed from manual ground-truth lesion masks and automaticSCIsegpredictions were correlated with clinical scores (pinprick, light touch, and lower extremity motor scores). A between-group comparison was performed using the Wilcoxon signed-rank test.ResultsMRI data from 191 SCI patients (mean age, 48.1 years ± 17.9 [SD]; 142 males) were used for training. Compared to existing methods,SCIsegachieved the best segmentation performance for both the cord and lesions and generalized well to both traumatic and non-traumatic SCI patients.SCIsegis open-source and accessible through the Spinal Cord Toolbox.ConclusionAutomatic segmentation of intramedullary lesions commonly seen in traumatic SCI replaces the tedious manual annotation process and enables the extraction of relevant lesion morphometrics in large cohorts. The proposed model generalizes across lesion etiologies (traumatic, ischemic), scanner manufacturers and heterogeneous image resolutions.SummaryAutomatic segmentation of the spinal cord and T2-weighted hyperintense lesions in spinal cord injury on MRI scans across different treatment strategies, lesion etiologies, sites, scanner manufacturers, and heterogeneous image resolutions.Key ResultsAn open-source, automatic method,SCIseg, was trained on a dataset of 191 spinal cord injury patients from three sites for the segmentation of spinal cord and T2-weighted hyperintense lesions.SCIseggeneralizes across traumatic and non-traumatic lesions, scanner manufacturers, and heterogeneous image resolutions, enabling the automatic extraction of lesion morphometrics in large multi-site cohorts.Morphometrics derived from the automatic predictions showed no statistically significant difference when compared with manual ground truth, implying reliability inSCIseg’spredictions.

show abstract

“…[ 24 ] Since that publication, scES’s ability to activate latent sensorimotor and autonomic networks to improve bowel, bladder, and cardiovascular regulation has been confirmed in a number of studies of small sample sizes. [ 5 , 11 , 14 , 21 , 25 , 48 ] Most recently, scES enabled over-ground walking with balance assistance in two previously motor complete SCI patients. [ 3 ] Gill and colleagues also describe independent stepping with scES after locomotor rehabilitation in an individual with complete sensorimotor lower extremity function loss.…”

Section: Introductionmentioning

confidence: 99%

Spinal cord epidural stimulation for motor and autonomic function recovery after chronic spinal cord injury: A case series and technical note

Boakye

Ball

Dietz

et al. 2023

Surgical Neurology International

Self Cite

View full text Add to dashboard Cite

Background: Traumatic spinal cord injury (tSCI) is a debilitating condition, leading to chronic morbidity and mortality. In recent peer-reviewed studies, spinal cord epidural stimulation (scES) enabled voluntary movement and return of over-ground walking in a small number of patients with motor complete SCI. Using the most extensive case series (n = 25) for chronic SCI, the present report describes our motor and cardiovascular and functional outcomes, surgical and training complication rates, quality of life (QOL) improvements, and patient satisfaction results after scES. Methods: This prospective study occurred at the University of Louisville from 2009 to 2020. scES interventions began 2–3 weeks after surgical implantation of the scES device. Perioperative complications were recorded as well as long-term complications during training and device related events. QOL outcomes and patient satisfaction were evaluated using the impairment domains model and a global patient satisfaction scale, respectively. Results: Twenty-five patients (80% male, mean age of 30.9 ± 9.4 years) with chronic motor complete tSCI underwent scES using an epidural paddle electrode and internal pulse generator. The interval from SCI to scES implantation was 5.9 ± 3.4 years. Two participants (8%) developed infections, and three additional patients required washouts (12%). All participants achieved voluntary movement after implantation. A total of 17 research participants (85%) reported that the procedure either met (n = 9) or exceeded (n = 8) their expectations, and 100% would undergo the operation again. Conclusion: scES in this series was safe and achieved numerous benefits on motor and cardiovascular regulation and improved patient-reported QOL in multiple domains, with a high degree of patient satisfaction. The multiple previously unreported benefits beyond improvements in motor function render scES a promising option for improving QOL after motor complete SCI. Further studies may quantify these other benefits and clarify scES’s role in SCI patients.

show abstract

Machine learning in clinical diagnosis, prognostication, and management of acute traumatic spinal cord injury (SCI): A systematic review

Cited by 23 publications

References 60 publications

Machine Learning Training Optimization using the Barycentric Correction Procedure

Machine Learning Training Optimization using the Barycentric Correction Procedure

SCIseg: Automatic Segmentation of T2-weighted Intramedullary Lesions in Spinal Cord Injury

Spinal cord epidural stimulation for motor and autonomic function recovery after chronic spinal cord injury: A case series and technical note

Contact Info

Product

Resources

About