Predicting task performance from upper extremity impairment measures after cervical spinal cord injury

Data-driven prediction of spinal cord injury recovery: An exploration of current status and future perspectives

Spinal Cord Injury (SCI) presents a significant challenge in rehabilitation medicine, with recovery outcomes varying widely among individuals. Machine learning (ML) is a promising approach to enhance the prediction of recovery trajectories, but its integration into clinical practice requires a thorough understanding of its efficacy and applicability. We systematically reviewed the current literature on data-driven models of SCI recovery prediction. The included studies were evaluated based on a range of criteria assessing the approach, implementation, input data preferences, and the clinical outcomes aimed to forecast. We observe a tendency to utilize routinely acquired data, such as International Standards for Neurological Classification of SCI (ISNCSCI), imaging, and demographics, for the prediction of functional outcomes derived from the Spinal Cord Independence Measure (SCIM) III and Functional Independence Measure (FIM) scores with a focus on motor ability. Although there has been an increasing interest in data-driven studies over time, traditional machine learning architectures, such as linear regression and tree-based approaches, remained the overwhelmingly popular choices for implementation. This implies ample opportunities for exploring architectures addressing the challenges of predicting SCI recovery, including techniques for learning from limited longitudinal data, improving generalizability, and enhancing reproducibility. We conclude with a perspective, highlighting possible future directions for data-driven SCI recovery prediction and drawing parallels to other application fields in terms of diverse data types (imaging, tabular, sequential, multimodal), data challenges (limited, missing, longitudinal data), and algorithmic needs (causal inference, robustness).

Predicting the Outcome and Survival of Patients with Spinal Cord Injury Using Machine Learning Algorithms: A Systematic Review

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.

Characterizing neurological status in individuals with tetraplegia using transcutaneous spinal stimulation

Transcutaneous spinal stimulation (TSS) is emerging as a valuable tool for electrophysiological and clinical assessment. This study had the objective of examining the recruitment patterns of upper limb (UL) motor pools through the delivery of TSS above and below a spinal lesion. It also aimed to explore the connection between the recruitment pattern of UL motor pools and the neurological and functional status following spinal cord injury (SCI). In eight participants with tetraplegia due to cervical SCI, TSS was delivered to the cervical spinal cord between the spinous processes of C3-C4 and C7-T1 vertebrae, and spinally evoked motor potentials in UL muscles were characterized. We found that responses observed in UL muscles innervated by motor pools below the level of injury demonstrated relatively reduced sensitivity to TSS compared to those above the lesion, were asymmetrical in the majority of muscles, and were dependent on the level, extent, and side of SCI. Overall, our findings indicate that electrophysiological data acquired through TSS can offer insights into the extent of UL functional asymmetry, disruptions in neural pathways, and changes in motor control following SCI. This study suggests that such electrophysiological data can supplement clinical and functional assessment and provide further insight regarding residual motor function in individuals with SCI.

Transcutaneous spinal stimulation provides characterization of neurological status in individuals with tetraplegia

Transcutaneous spinal stimulation (TSS) is emerging as a valuable tool for electrophysiological and clinical assessment. This study had the objective of examining the recruitment patterns of upper limb (UL) motor pools through the delivery of TSS above and below a spinal lesion. It also aimed to explore the connection between the recruitment pattern of UL motor pools and the neurological and functional status following spinal cord injury (SCI). In eight participants with tetraplegia due to cervical SCI, TSS was delivered to the cervical spinal cord between the spinous processes of C3-C4 and C7-T1 vertebrae, and spinally evoked motor potentials in UL muscles were characterized. We found that responses observed in UL muscles innervated by motor pools below the level of injury demonstrated relatively reduced sensitivity to TSS compared to those above the lesion, were asymmetrical in the majority of muscles, and were dependent on the level, extent, and side of SCI. Overall, our findings indicate that electrophysiological data acquired through TSS can offer insights into the extent of UL functional asymmetry, disruptions in neural pathways, and changes in motor control following SCI. This study suggests that such electrophysiological data can supplement clinical and functional assessment and provide further insight regarding residual motor function in individuals with SCI.

The Use of Artificial Intelligence to Predict the Prognosis of Patients Undergoing Central Nervous System Rehabilitation: A Narrative Review

Applications of machine learning in the healthcare field have become increasingly diverse. In this review, we investigated the integration of artificial intelligence (AI) in predicting the prognosis of patients with central nervous system disorders such as stroke, traumatic brain injury, and spinal cord injury. AI algorithms have shown promise in prognostic assessment, but challenges remain in achieving a higher prediction accuracy for practical clinical use. We suggest that accumulating more diverse data, including medical imaging and collaborative efforts among hospitals, can enhance the predictive capabilities of AI. As healthcare professionals become more familiar with AI, its role in central nervous system rehabilitation is expected to advance significantly, revolutionizing patient care.

Functional electrical stimulation therapy for upper extremity rehabilitation following spinal cord injury: a pilot study

STUDY DESIGN

Pilot study.

OBJECTIVES

To examine if functional electrical stimulation therapy (FEST) improves neuromuscular factors underlying upper limb function in individuals with SCI.

SETTING

A tertiary spinal cord rehabilitation center specialized in spinal cord injury care in Canada.

METHODS

We examined 29 muscles from 4 individuals living with chronic, cervical, and incomplete SCI. The analysis was focused on the changes in muscle activation, as well as on how the treatment could change the ability to control a given muscle or on how multiple muscles would be coordinated during volitional efforts.

RESULTS

There was evidence of gains in muscle strength, activation, and median frequency after the FEST. Gains in muscle activation indicated the activation of a greater number of motor units and gains in muscle median frequency the involvement of higher threshold, faster motor units. In some individuals, these changes were smaller but accompanied by increased control over muscle contraction, evident in a greater ability to sustain a volitional contraction, reduce the co-contraction of antagonist muscles, and provide cortical drive.

CONCLUSIONS

FEST increases muscle strength and activation. Enhanced control of muscle contraction, reduced co-contraction of antagonist muscles, and a greater presence of cortical drive were some of the findings supporting the effects of FEST at the sensory-motor integration level.

Segmental motor recovery after cervical spinal cord injury relates to density and integrity of corticospinal tract projections

Cervical spinal cord injury (SCI) causes extensive impairments for individuals which may include dextrous hand function. Although prior work has focused on the recovery at the person-level, the factors determining the recovery of individual muscles are poorly understood. Here, we investigate the muscle-specific recovery after cervical spinal cord injury in a retrospective analysis of 748 individuals from the European Multicenter Study about Spinal Cord Injury (NCT01571531). We show associations between corticospinal tract (CST) sparing and upper extremity recovery in SCI, which improves the prediction of hand muscle strength recovery. Our findings suggest that assessment strategies for muscle-specific motor recovery in acute spinal cord injury are improved by accounting for CST sparing, and complement person-level predictions.

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

Background

Machine learning has been applied to improve diagnosis and prognostication of acute traumatic spinal cord injury. We investigate potential for clinical integration of machine learning in this patient population to navigate variability in injury and recovery.

Materials and methods

We performed a systematic review using PRISMA guidelines through PubMed database to identify studies that use machine learning algorithms for clinical application toward improvements in diagnosis, management, and predictive modeling.

Results

Of the 132 records identified, a total of 13 articles met inclusion criteria and were included in final analysis. Of the 13 articles, 5 focused on diagnostic accuracy and 8 were related to prognostication or management of traumatic spinal cord injury. Across studies, 1983 patients with spinal cord injury were evaluated with most classifying as ASIA C or D. Retrospective designs were used in 10 of 13 studies and 3 were prospective. Studies focused on MRI evaluation and segmentation for diagnostic accuracy and prognostication, investigation of mean arterial pressure in acute care and intraoperative settings, prediction of ambulatory and functional ability, chronic complication prevention, and psychological quality of life assessments. Decision tree, random forests (RF), support vector machines (SVM), hierarchical cluster tree analysis (HCTA), artificial neural networks (ANN), convolutional neural networks (CNN) machine learning subtypes were used.

Conclusions

Machine learning represents a platform technology with clinical application in traumatic spinal cord injury diagnosis, prognostication, management, rehabilitation, and risk prevention of chronic complications and mental illness. SVM models showed improved accuracy when compared to other ML subtypes surveyed. Inherent variability across patients with SCI offers unique opportunity for ML and personalized medicine to drive desired outcomes and assess risks in this patient population.

Development of Reaching, Grasping & Manipulation indicators to advance the quality of spinal cord injury rehabilitation: SCI-High Project

OBJECTIVE

To describe the development of structure, process, and outcome indicators aimed to advance the quality of Reaching, Grasping & Manipulation (RG&M) rehabilitation for Canadians living with spinal cord injury or disease (SCI/D).

METHOD

Upper extremity rehabilitation experts developed a framework of indicators for evaluation of RG&M rehabilitation quality. A systematic search of the literature identified potential upper extremity indicators that influence RG&M outcomes. A Driver diagram summarized factors influencing upper extremity outcomes to inform the selection of structure and process indicators. Psychometric properties, clinical utility, and feasibility of potential upper extremity measures were considered when selecting outcome indicators.

RESULTS

The selected structure indicator is the number of occupational and physical therapists with specialized certification, education, training and/or work experience in upper extremity therapy related to RG&M at a given SCI/D rehabilitation center. The process indicator is the total hours of upper extremity therapies related to RG&M and the proportion of this time allocated to neurorestorative therapy for each individual with tetraplegia receiving therapy. The outcome indicators are the Graded Redefined Assessment of Strength, Sensation and Prehension (GRASSP) strength and Spinal Cord Independence Measure III (SCIM III) Self-Care subscores implemented at rehabilitation admission and discharge, and SCIM III Self-Care subscore only at 18 months post-admission.

CONCLUSION

The selected indicators align with current practice, will direct the timing of routine assessments, and enhance the volume and quality of RG&M therapy delivered, with the aim to ultimately increase the proportion of individuals with tetraplegia achieving improved upper extremity function by 18 months post-rehabilitation.

Effects of Combined Upper Limb Robotic Therapy in Patients With Tetraplegic Spinal Cord Injury

OBJECTIVE

To confirm the effects of combined upper limb robotic therapy (RT) as compared to conventional occupational therapy (OT) in tetraplegic spinal cord injury (SCI) patients and to suggest the optimized treatment guidelines of combined upper limb RT.

METHODS

After subject recruitment and screening for eligibility, the baseline evaluation for outcome measures were performed. We evaluated the Graded and Redefined Assessment of Strength, Sensibility, and Prehension (GRASSP), the American Spinal Injury Association upper extremity motor score, grip and pinch strength, and the Spinal Cord Independence Measurement III (SCIM-III). In this study, the pre-tested participants were divided randomly into the RT and OT group. The utilized interventions included combined upper limb RT using ArmeoPower and Amadeo (RT group), or conventional OT (OT group) in addition to daily inpatient rehabilitation program. The participants underwent 40 minutes×3 sessions×5 weeks of interventions.

RESULTS

A total of 30 tetraplegic SCI patients completed entire study program. After 5 weeks of intervention, both groups demonstrated increases in GRASSP-strength and SCIM-III. The manual muscle test scores of elbow flexion, elbow extension, 2-5th metacarpophalangeal extension, and SCIM-III subscores of bathing-upper, dressing-upper, and grooming as well as the GRASSP-qualitative prehension score were noted to have been significantly increased in the RT group as evaluated. The OT group showed improvements in the GRASSP-quantitative prehension score and some items in grip and pinch strength. There was no significant difference between the two groups in almost all measurements except for the SCIM-III bathing-upper subscore.

CONCLUSION

Combined upper limb RT demonstrated beneficial effects on the upper limb motor function in patients with tetraplegic SCI, which were comparable with conventional OT.

Clinical Outcomes from a Multi-Center Study of Human Neural Stem Cell Transplantation in Chronic Cervical Spinal Cord Injury

Human neural stem cell transplantation (HuCNS-SC^®) is a promising central nervous system (CNS) tissue repair strategy in patients with stable neurological deficits from chronic spinal cord injury (SCI). These immature human neural cells have been demonstrated to survive when transplanted in vivo, extend neural processes, form synaptic contacts, and improve functional outcomes after experimental SCI. A phase II single blind, randomized proof-of-concept study of the safety and efficacy of HuCNS-SC transplantation into the cervical spinal cord was undertaken in patients with chronic C5-7 tetraplegia, 4-24 months post-injury. In Cohort I (n = 6) dose escalation from 15,000,000 to 40,000,000 cells was performed to determine the optimum dose. In Cohort II an additional six participants were transplanted at target dose (40,000,000) and compared with four untreated controls. Within the transplant group, there were nine American Spinal Injury Association Impairment Scale (AIS) B and three AIS A participants with a median age at transplant of 28 years with an average time to transplant post-injury of 1 year. Immunosuppression was continued for 6 months post-transplant, and immunosuppressive blood levels of tacrolimus were achieved and well tolerated. At 1 year post-transplantation, there was no evidence of additional spinal cord damage, new lesions, or syrinx formation on magnetic resonance (MR) imaging. In summary, the incremental dose escalation design established surgical safety, tolerability, and feasibility in Cohort I. Interim analysis of Cohorts I and II demonstrated a trend toward Upper Extremity Motor Score (UEMS) and Graded Redefined Assessment of Strength, Sensibility, and Prehension (GRASSP) motor gains in the treated participants, but at a magnitude below the required clinical efficacy threshold set by the sponsor to support further development resulting in early study termination.