An Interpretable Machine Learning Approach to Predict Survival Outcomes in Spinal and Sacropelvic Chordomas

STUDY DESIGN

Retrospective, population-based cohort study.

OBJECTIVE

This study aimed to develop machine learning (ML) models to predict five-year and 10-year mortality in spinal and sacropelvic chordoma patients and integrate them into a web application for enhanced prognostication.

SUMMARY OF BACKGROUND DATA

Past research has uncovered factors influencing survival in spinal chordoma patients. While identifying individual predictors is important, personalized survival predictions are equally vital. Though prior efforts have resulted in nomograms aiming to serve this purpose, they cannot capture complex interactions within data and rely on statistical assumptions that may not fit real-world data.

METHODS

Adult spinal and sacropelvic chordoma patients were identified from the National Cancer Database. Sociodemographic, clinicopathologic, diagnostic, and treatment-related variables were utilized as predictive features. Five supervised ML algorithms (TabPFN, CatBoost, XGBoost, LightGBM, and Random Forest) were implemented to predict mortality at five and 10 years postdiagnosis. Model performance was primarily evaluated using the area under the receiver operating characteristic (AUROC). SHapley Additive exPlanations (SHAP) values and partial dependence plots provided feature importance and interpretability. The top models were integrated into a web application.

RESULTS

From the NCDB, 1206 adult patients diagnosed with histologically confirmed spinal and sacropelvic chordomas were retrieved for the five-year mortality outcome [423 (35.1%) with five-year mortality] and 801 patients for the 10-year mortality outcome [588 (73.4%) with 10-year mortality]. Top-performing models for both of the outcomes were the models created with the CatBoost algorithm. The CatBoost model for five-year mortality predictions displayed a mean AUROC of 0.801, and the CatBoost model predicting 10-year mortality yielded a mean AUROC of 0.814.

CONCLUSIONS

This study developed ML models that can accurately predict five-year to 10-year survival probabilities in spinal chordoma patients. Integrating these interpretable, personalized prognostic models into a web application provides quantitative survival estimates for a given patient. The local interpretability enables transparency into how predictions are influenced. Further external validation is warranted to support generalizability and clinical utility.

Tracing topics and trends in drug-resistant epilepsy research using a natural language processing-based topic modeling approach

Epilepsy is a common neurological disorder affecting over 70 million people worldwide. Although many patients achieve seizure control with anti-epileptic drugs (AEDs), 30%-40% develop drug-resistant epilepsy (DRE), where seizures persist despite adequate trials of AEDs. DRE is associated with reduced quality of life, increased mortality and morbidity, and greater socioeconomic challenges. The continued intractability of DRE has fueled exponential growth in research that aims to understand and treat this serious condition. However, synthesizing this vast and continuously expanding DRE literature to derive insights poses considerable difficulties for investigators and clinicians. Conventional review methods are often prolonged, hampering the timely application of findings. More-efficient approaches to analyze the voluminous research are needed. In this study, we utilize a natural language processing (NLP)-based topic modeling approach to examine the DRE publication landscape, uncovering key topics and trends. Documents were retrieved from Scopus, preprocessed, and modeled using BERTopic. This technique employs transformer models like BERT (Bidirectional Encoder Representations from Transformers) for contextual understanding, thereby enabling accurate topic categorization. Analysis revealed 18 distinct topics spanning various DRE research areas. The 10 most common topics, including "AEDs," "Neuromodulation Therapy," and "Genomics," were examined further. "Cannabidiol," "Functional Brain Mapping," and "Autoimmune Encephalitis" emerged as the hottest topics of the current decade, and were examined further. This NLP methodology provided valuable insights into the evolving DRE research landscape, revealing shifting priorities and declining interests. Moreover, we demonstrate an efficient approach to synthesizing and visualizing patterns within extensive literature that could be applied to other research fields.

From Text to Insight: A Natural Language Processing-Based Analysis of Topics and Trends in Neurosurgery

BACKGROUND AND OBJECTIVES

Neurosurgical research is a rapidly evolving field, with new research topics emerging continually. To provide a clearer understanding of the evolving research landscape, our study aimed to identify and analyze the prevalent research topics and trends in Neurosurgery.

METHODS

We used BERTopic, an advanced natural language processing-based topic modeling approach, to analyze papers published in the journal Neurosurgery . Using this method, topics were identified based on unique sets of keywords that encapsulated the core themes of each article. Linear regression models were then trained on the topic probabilities to identify trends over time, allowing us to identify "hot" (growing in prominence) and "cold" (decreasing in prominence) topics. We also performed a focused analysis of the trends in the current decade.

RESULTS

Our analysis led to the categorization of 12 438 documents into 49 distinct topics. The topics covered a wide range of themes, with the most commonly identified topics being "Spinal Neurosurgery" and "Treatment of Cerebral Ischemia." The hottest topics of the current decade were "Peripheral Nerve Surgery," "Unruptured Aneurysms," and "Endovascular Treatments" while the cold topics were "Chiari Malformations," "Thromboembolism Prophylaxis," and "Infections."

CONCLUSION

Our study underscores the dynamic nature of neurosurgical research and the evolving focus of the field. The insights derived from the analysis can guide future research directions, inform policy decisions, and identify emerging areas of interest. The use of natural language processing in synthesizing and analyzing large volumes of academic literature demonstrates the potential of advanced analytical techniques in understanding the research landscape, paving the way for similar analyses across other medical disciplines.

Exploiting Natural Language Processing to Unveil Topics and Trends of Traumatic Brain Injury Research

Traumatic brain injury (TBI) has evolved from a topic of relative obscurity to one of widespread scientific and lay interest. The scope and focus of TBI research have shifted, and research trends have changed in response to public and scientific interest. This study has two primary goals: first, to identify the predominant themes in TBI research; and second, to delineate "hot" and "cold" areas of interest by evaluating the current popularity or decline of these topics. Hot topics may be dwarfed in absolute numbers by other, larger TBI research areas but are rapidly gaining interest. Likewise, cold topics may present opportunities for researchers to revisit unanswered questions. We utilized BERTopic, an advanced natural language processing (NLP)-based technique, to analyze TBI research articles published since 1990. This approach facilitated the identification of key topics by extracting sets of distinctive keywords representative of each article's core themes. Using these topics' probabilities, we trained linear regression models to detect trends over time, recognizing topics that were gaining (hot) or losing (cold) relevance. Additionally, we conducted a specific analysis focusing on the trends observed in TBI research in the current decade (the 2020s). Our topic modeling analysis categorized 42,422 articles into 27 distinct topics. The 10 most frequently occurring topics were: "Rehabilitation," "Molecular Mechanisms of TBI," "Concussion," "Repetitive Head Impacts," "Surgical Interventions," "Biomarkers," "Intracranial Pressure," "Posttraumatic Neurodegeneration," "Chronic Traumatic Encephalopathy," and "Blast Induced TBI," while our trend analysis indicated that the hottest topics of the current decade were "Genomics," "Sex Hormones," and "Diffusion Tensor Imaging," while the cooling topics were "Posttraumatic Sleep," "Sensory Functions," and "Hyperosmolar Therapies." This study highlights the dynamic nature of TBI research and underscores the shifting emphasis within the field. The findings from our analysis can aid in the identification of emerging topics of interest and areas where there is little new research reported. By utilizing NLP to effectively synthesize and analyze an extensive collection of TBI-related scholarly literature, we demonstrate the potential of machine learning techniques in understanding and guiding future research prospects. This approach sets the stage for similar analyses in other medical disciplines, offering profound insights and opportunities for further exploration.

Scoping Review with Topic Modeling on the Diagnostic Criteria for Degenerative Cervical Myelopathy

STUDY DESIGN

This study is a scoping review.

OBJECTIVE

There is a broad variability in the definition of degenerative cervical myelopathy (DCM) and no standardized set of diagnostic criteria to date.

METHODS

We interrogated the Myelopathy.org database, a hand-indexed database of primary clinical studies conducted exclusively on DCM in humans between 2005-2021. The DCM inclusion criteria used in these studies were inputted into 3 topic modeling algorithms: Hierarchical Dirichlet Process (HDP), Latent Dirichlet Allocation (LDA), and BERtopic. The emerging topics were subjected to manual labeling and interpretation.

RESULTS

Of 1676 reports, 120 papers (7.16%) had well-defined inclusion criteria and were subjected to topic modeling. Four topics emerged from the HDP model: disturbance from extremity weakness and motor signs; fine-motor and sensory disturbance of upper extremity; a combination of imaging and clinical findings is required for the diagnosis; and "reinforcing" (or modifying) factors that can aid in the diagnosis in borderline cases. The LDA model showed the following topics: disturbance to the patient is required for the diagnosis; reinforcing factors can aid in the diagnosis in borderline cases; clinical findings from the extremities; and a combination of imaging and clinical findings is required for the diagnosis. BERTopic identified the following topics: imaging abnormality, typical clinical features, range of objective criteria, and presence of clinical findings.

CONCLUSIONS

This review provides quantifiable data that only a minority of past studies in DCM provided meaningful inclusion criteria. The items and patterns found here are very useful for the development of diagnostic criteria for DCM.

Utilizing imaging parameters for functional outcome prediction in acute ischemic stroke: A machine learning study

BACKGROUND AND PURPOSE

We aimed to predict the functional outcome of acute ischemic stroke patients with anterior circulation large vessel occlusions (LVOs), irrespective of how they were treated or the severity of the stroke at admission, by only using imaging parameters in machine learning models.

METHODS

Consecutive adult patients with anterior circulation LVOs who were scanned with CT angiography (CTA) and CT perfusion were queried in this single-center, retrospective study. The favorable outcome was defined as a modified Rankin score (mRS) of 0-2 at 90 days. Predictor variables included only imaging parameters. CatBoost, XGBoost, and Random Forest were employed. Algorithms were evaluated using the area under the receiver operating characteristic curve (AUROC), the area under the precision-recall curve (AUPRC), accuracy, Brier score, recall, and precision. SHapley Additive exPlanations were implemented.

RESULTS

A total of 180 patients (102 female) were included, with a median age of 69.5. Ninety-two patients had an mRS between 0 and 2. The best algorithm in terms of AUROC was XGBoost (0.91). Furthermore, the XGBoost model exhibited a precision of 0.72, a recall of 0.81, an AUPRC of 0.83, an accuracy of 0.78, and a Brier score of 0.17. Multiphase CTA collateral score was the most significant feature in predicting the outcome.

CONCLUSIONS

Using only imaging parameters, our model had an AUROC of 0.91 which was superior to most previous studies, indicating that imaging parameters may be as accurate as conventional predictors. The multiphase CTA collateral score was the most predictive variable, highlighting the importance of collaterals.

Predicting 30-Day Non-Seizure Outcomes Following Temporal Lobectomy with Personalized Machine Learning Models

BACKGROUND

Temporal lobe epilepsy is the most common reason behind drug-resistant seizures and temporal lobectomy (TL) is performed after all other efforts have been taken for a Temporal lobe epilepsy. Our study aims to develop multiple machine learning (ML) models capable of predicting postoperative outcomes following TL surgery.

METHODS

Data from the American College of Surgeons National Surgical Quality Improvement Program database identified patients who underwent TL surgery. We focused on 3 outcomes: prolonged length of stay (LOS), nonhome discharges, and 30-day readmissions. Six ML algorithms, TabPFN, XGBoost, LightGBM, Support Vector Machine, Random Forest, and Logistic Regression, coupled with the Optuna optimization library for hyperparameter tuning, were tested. Models with the highest area under the receiver operating characteristic (AUROC) values were included in the web application. SHapley Additive exPlanations was used to evaluate importance of predictor variables.

RESULTS

Our analysis included 423 patients. Of these patients, 111 (26.2%) experienced prolonged LOS, 33 (7.8%) had nonhome discharges, and 29 (6.9%) encountered 30-day readmissions. The top-performing models for each outcome were those built with the Random Forest algorithm. The Random Forest models yielded AUROCs of 0.868, 0.804, and 0.742 in predicting prolonged LOS, nonhome discharges, and 30-day readmissions, respectively.

CONCLUSIONS

Our study uses ML to forecast adverse postoperative outcomes following TL. We developed accessible predictive models that enhance prognosis prediction for TL surgery. Making ML models available for this purpose represents a significant advancement in shifting toward a more patient-centric, data-driven paradigm.

Natural language processing reveals research trends and topics in The Spine Journal over two decades: a topic modeling study

BACKGROUND CONTEXT

The field of spine research is rapidly evolving, with new research topics continually emerging. Analyzing topics and trends in the literature can provide insights into the shifting research landscape.

PURPOSE

This study aimed to elucidate prevalent and emerging research topics and trends within The Spine Journal using a natural language processing technique called topic modeling.

METHODS

We utilized BERTopic, a topic modeling technique rooted in natural language processing (NLP), to examine articles from The Spine Journal. Through this approach, we discerned topics from distinct keyword clusters and representative documents that represented the main concepts of each topic. We then used linear regression models on these topic likelihoods to trace trends over time, pinpointing both "hot" (growing in prominence) and "cold" (decreasing in prominence) topics. Additionally, we conducted an in-depth review of the trending topics in the present decade.

RESULTS

Our analysis led to the categorization of 3358 documents into 30 distinct topics. These topics spanned a wide range of themes, with the most commonly identified topics being "Outcome Measures," "Scoliosis," and "Intradural Lesions." Throughout the history of the journal, the three hottest topics were "Degenerative Cervical Myelopathy," "Osteoporosis," and "Opioid Use." Conversely, the coldest topics were "Intradural Lesions," "Extradural Tumors," and "Vertebral Augmentation." Within the current decade, the hottest topics were "Screw Biomechanics," "Paraspinal Muscles," and "Biologics for Fusion," whereas the cold topics were "Intraoperative Blood Loss," "Construct Biomechanics," and "Material Science."

CONCLUSIONS

This study accentuates the dynamic nature of spine research and the changing focus within the field. The insights gleaned from our analysis can steer future research directions, inform policy decisions, and spotlight emerging areas of interest. The implementation of NLP to synthesize and analyze vast amounts of academic literature exhibits the potential of advanced analytical techniques in comprehending the research landscape, setting a precedent for similar analyses across other medical disciplines.

Cervical Spine Stenosis Causing Diaphragmatic Paralysis: A Case Study and Narrative Review of Clinical Presentations and Outcomes

STUDY DESIGN

Case report and narrative review.

OBJECTIVE

To explore the therapeutic role of surgical and nonsurgical treatment of diaphragmatic paralysis secondary to spinal cord and nerve root compression.

SUMMARY OF BACKGROUND DATA

Phrenic nerve dysfunction due to central or neuroforaminal stenosis is a rare yet unappreciated etiology of diaphragmatic paralysis and chronic dyspnea. Surgical spine decompression, diaphragmatic pacing, and intensive physiotherapy are potential treatment options with varying degrees of evidence.

METHODS

The case of a 70-year-old male with progressive dyspnea, reduced hemi-diaphragmatic excursion, and C3-C7 stenosis, who underwent a microscopic foraminotomy is discussed. Literature review (MEDLINE, PubMed, Google Scholar) identified 19 similar reports and discussed alternative treatments and outcomes.

RESULTS AND CONCLUSIONS

Phrenic nerve root decompression and improvement in neuromonitoring signals were observed intraoperatively. The patient's postoperative course was uncomplicated, and after 15 months, he experienced significant symptomatic improvement and minor improvement in hemi-diaphragmatic paralysis and pulmonary function tests. All case reports of patients treated with spinal decompression showed symptomatic and/or functional improvement, while one of the 2 patients treated with physiotherapy showed improvement. More studies are needed to further describe the course and outcomes of these interventions, but early identification and spinal decompression can be an effective treatment.

OCEBM LEVEL OF EVIDENCE

Level-4.

Trends in stroke-related journals: Examination of publication patterns using topic modeling

OBJECTIVES

This study aims to demonstrate the capacity of natural language processing and topic modeling to manage and interpret the vast quantities of scholarly publications in the landscape of stroke research. These tools can expedite the literature review process, reveal hidden themes, and track rising research areas.

MATERIALS AND METHODS

Our study involved reviewing and analyzing articles published in five prestigious stroke journals, namely Stroke, International Journal of Stroke, European Stroke Journal, Translational Stroke Research, and Journal of Stroke and Cerebrovascular Diseases. The team extracted document titles, abstracts, publication years, and citation counts from the Scopus database. BERTopic was chosen as the topic modeling technique. Using linear regression models, current stroke research trends were identified. Python 3.1 was used to analyze and visualize data.

RESULTS

Out of the 35,779 documents collected, 26,732 were classified into 30 categories and used for analysis. "Animal Models," "Rehabilitation," and "Reperfusion Therapy" were identified as the three most prevalent topics. Linear regression models identified "Emboli," "Medullary and Cerebellar Infarcts," and "Glucose Metabolism" as trending topics, whereas "Cerebral Venous Thrombosis," "Statins," and "Intracerebral Hemorrhage" demonstrated a weaker trend.

CONCLUSIONS

The methodology can assist researchers, funders, and publishers by documenting the evolution and specialization of topics. The findings illustrate the significance of animal models, the expansion of rehabilitation research, and the centrality of reperfusion therapy. Limitations include a five-journal cap and a reliance on high-quality metadata.

Development and internal validation of machine learning models for personalized survival predictions in spinal cord glioma patients

BACKGROUND CONTEXT

Numerous factors have been associated with the survival outcomes in patients with spinal cord gliomas (SCG). Recognizing these specific determinants is crucial, yet it is also vital to establish a reliable and precise prognostic model for estimating individual survival outcomes.

OBJECTIVE

The objectives of this study are twofold: first, to create an array of interpretable machine learning (ML) models developed for predicting survival outcomes among SCG patients; and second, to integrate these models into an easily navigable online calculator to showcase their prospective clinical applicability.

STUDY DESIGN

This was a retrospective, population-based cohort study aiming to predict the outcomes of interest, which were binary categorical variables, in SCG patients with ML models.

PATIENT SAMPLE

The National Cancer Database (NCDB) was utilized to identify adults aged 18 years or older who were diagnosed with histologically confirmed SCGs between 2010 and 2019.

OUTCOME MEASURES

The outcomes of interest were survival outcomes at three specific time points post-diagnosis: 1, 3, and 5 years. These outcomes were formed by combining the "Vital Status" and "Last Contact or Death (Months from Diagnosis)" variables. Model performance was evaluated visually and numerically. The visual evaluation utilized receiver operating characteristic (ROC) curves, precision-recall curves (PRCs), and calibration curves. The numerical evaluation involved metrics such as sensitivity, specificity, accuracy, area under the PRC (AUPRC), area under the ROC curve (AUROC), and Brier Score.

METHODS

We employed five ML algorithms-TabPFN, CatBoost, XGBoost, LightGBM, and Random Forest-along with the Optuna library for hyperparameter optimization. The models that yielded the highest AUROC values were chosen for integration into the online calculator. To enhance the explicability of our models, we utilized SHapley Additive exPlanations (SHAP) for assessing the relative significance of predictor variables and incorporated partial dependence plots (PDPs) to delineate the influence of singular variables on the predictions made by the top performing models.

RESULTS

For the 1-year survival analysis, 4,913 patients [5.6% with 1-year mortality]; for the 3-year survival analysis, 4,027 patients (11.5% with 3-year mortality]; and for the 5-year survival analysis, 2,854 patients (20.4% with 5-year mortality) were included. The top models achieved AUROCs of 0.938 for 1-year mortality (TabPFN), 0.907 for 3-year mortality (LightGBM), and 0.902 for 5-year mortality (Random Forest). Global SHAP analyses across survival outcomes at different time points identified histology, tumor grade, age, surgery, radiotherapy, and tumor size as the most significant predictor variables for the top-performing models.

CONCLUSIONS

This study demonstrates ML techniques can develop highly accurate prognostic models for SCG patients with excellent discriminatory ability. The interactive online calculator provides a tool for assessment by physicians (https://huggingface.co/spaces/MSHS-Neurosurgery-Research/NCDB-SCG). Local interpretability informs prediction influences for a given individual. External validation across diverse datasets could further substantiate potential utility and generalizability. This robust, interpretable methodology aligns with the goals of precision medicine, establishing a foundation for continued research leveraging ML's predictive power to enhance patient counseling.

Ultra-Early (<5 Hours) Decompression for Thoracolumbar Spinal Cord Injury: A Case Series

Early surgical decompression within 24 hours for traumatic spinal cord injury (SCI) is associated with improved neurological recovery. However, the ideal timing of decompression is still up for debate. The objective of this study was to utilize our retrospective single-institution series of ultra-early (<5 hours) decompression to determine if ultra-early decompression led to improved neurological outcomes and was a feasible target over previously defined early decompression targets. Retrospective data on patients with SCI who underwent ultra-early (<5 hours) decompression at a level one metropolitan trauma center were extracted and collected from 2015-2018. American Spinal Injury Association (ASIA) Impairment Scale (AIS) grade improvement was the primary outcome, with ASIA Motor score improvement and complication rate as secondary outcomes. Four individuals met the criteria for inclusion in this case series. All four suffered thoracolumbar SCI. All patients improved neurologically by AIS grade, and there were no complications directly related to ultra-early surgery. Given the small sample size, there was no statistically significant difference in outcomes compared to a control group who underwent early (5-24 hour) decompression in the same period. Ultra-early decompression is a feasible and safe target for thoracolumbar SCI and may lead to improved neurological outcomes without increased risk of complications. This case series can help create the foundation for future, larger studies that may definitively show the benefit of ultra-early decompression.

The Predictive Abilities of Machine Learning Algorithms in Patients with Thoracolumbar Spinal Cord Injuries

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.

Personalized Prognosis with Machine Learning Models for Predicting In-Hospital Outcomes Following Intracranial Meningioma Resections

BACKGROUND

Meningiomas display diverse biological traits and clinical behaviors, complicating patient outcome prediction. This heterogeneity, along with varying prognoses, underscores the need for a precise, personalized evaluation of postoperative outcomes.

METHODS

Data from the American College of Surgeons National Surgical Quality Improvement Program database identified patients who underwent intracranial meningioma resections from 2014 to 2020. We focused on 5 outcomes: prolonged LOS, nonhome discharges, 30-day readmissions, unplanned reoperations, and major complications. Six machine learning algorithms, including TabPFN, TabNet, XGBoost, LightGBM, Random Forest, and Logistic Regression, coupled with the Optuna optimization library for hyperparameter tuning, were tested. Models with the highest area under the receiver operating characteristic (AUROC) values were included in the web application. SHapley Additive exPlanations were used to evaluate the importance of predictor variables.

RESULTS

Our analysis included 7000 patients. Of these patients, 1658 (23.7%) had prolonged LOS, 1266 (18.1%) had nonhome discharges, 573 (8.2%) had 30-day readmission, 253 (3.6%) had unplanned reoperation, and 888 (12.7%) had major complications. Performance evaluation indicated that the top-performing models for each outcome were the models built with LightGBM and Random Forest algorithms. The LightGBM models yielded AUROCs of 0.842 and 0.846 in predicting prolonged LOS and nonhome discharges, respectively. The Random Forest models yielded AUROCs of 0.717, 0.76, and 0.805 in predicting 30-day readmissions, unplanned reoperations, and major complications, respectively.

CONCLUSIONS

The study successfully demonstrated the potential of machine learning models in predicting short-term adverse postoperative outcomes after meningioma resections. This approach represents a significant step forward in personalizing the information provided to meningioma patients.

Machine Learning and Symptom Patterns in Degenerative Cervical Myelopathy: Web-Based Survey Study

BACKGROUND

Degenerative cervical myelopathy (DCM), a progressive spinal cord injury caused by spinal cord compression from degenerative pathology, often presents with neck pain, sensorimotor dysfunction in the upper or lower limbs, gait disturbance, and bladder or bowel dysfunction. Its symptomatology is very heterogeneous, making early detection as well as the measurement or understanding of the underlying factors and their consequences challenging. Increasingly, evidence suggests that DCM may consist of subgroups of the disease, which are yet to be defined.

OBJECTIVE

This study aimed to explore whether machine learning can identify clinically meaningful groups of patients based solely on clinical features.

METHODS

A survey was conducted wherein participants were asked to specify the clinical features they had experienced, their principal presenting complaint, and time to diagnosis as well as demographic information, including disease severity, age, and sex. K-means clustering was used to divide respondents into clusters according to their clinical features using the Euclidean distance measure and the Hartigan-Wong algorithm. The clinical significance of groups was subsequently explored by comparing their time to presentation, time with disease severity, and other demographics.

RESULTS

After a review of both ancillary and cluster data, it was determined by consensus that the optimal number of DCM response groups was 3. In Cluster 1, there were 40 respondents, and the ratio of male to female participants was 13:21. In Cluster 2, there were 92 respondents, with a male to female participant ratio of 27:65. Cluster 3 had 57 respondents, with a male to female participant ratio of 9:48. A total of 6 people did not report biological sex in Cluster 1. The mean age in this Cluster was 56.2 (SD 10.5) years; in Cluster 2, it was 54.7 (SD 9.63) years; and in Cluster 3, it was 51.8 (SD 8.4) years. Patients across clusters significantly differed in the total number of clinical features reported, with more clinical features in Cluster 3 and the least clinical features in Cluster 1 (Kruskal-Wallis rank sum test: χ22=159.46; P<.001). There was no relationship between the pattern of clinical features and severity. There were also no differences between clusters regarding time since diagnosis and time with DCM.

CONCLUSIONS

Using machine learning and patient-reported experience, 3 groups of patients with DCM were defined, which were different in the number of clinical features but not in the severity of DCM or time with DCM. Although a clearer biological basis for the clusters may have been missed, the findings are consistent with the emerging observation that DCM is a heterogeneous disease, difficult to diagnose or stratify. There is a place for machine learning methods to efficiently assist with pattern recognition. However, the challenge lies in creating quality data sets necessary to derive benefit from such approaches.

Exploring the differences in radiologic and clinical outcomes of transforaminal lumbar interbody fusion with single- and bi-planar expandable cages: a systematic review and meta-analysis

Transforaminal lumbar interbody fusion (TLIF) is a universal surgical technique used to achieve lumbar fusion. Traditionally static cages have been used to restore the disc space after discectomy. However, newer technological advancements have brought up uniplanar expandable cages (UECs) and more recently bi-planar expandable cages (BECs), the latter with the hope of reducing the events of intra- or postoperative subsidence compared to UECs. However, since BECs are relatively new, there has been no comparison to UECs. In this PRISMA-compliant systematic review, we sought to identify all Medline and Embase reports that used UECs and/or BECs for TLIF or posterior lumbar interbody fusion. Primary outcomes included subsidence and fusion rates. Secondary outcomes included VAS back pain score, VAS leg pain score, ODI, and other complications. A meta-analysis of proportions was the main method used to evaluate the extracted data. Bias was assessed using the ROBINS-I tool. A total of 15 studies were pooled in the analysis, 3 of which described BECs. There were no studies directly comparing the UECs to BECs. A statistically significant difference in fusion rates was found between UECs and BECs (p = 0.04). Due to lack of direct comparative literature, definitive conclusions cannot be made about differences between UECs and BECs. The analysis showed a statistically higher fusion rate for BECs versus UECs, but this should be interpreted cautiously. No other statistically significant differences were found. As more direct comparative studies emerge, future meta-analyses may clarify potential differences between these cage types.

Comparative Effectiveness of Open Versus Minimally Invasive Transforaminal Lumbar Interbody Fusion: An Umbrella Review of Meta-Analyses

STUDY DESIGN

Umbrella review of meta-analyses.

OBJECTIVE

To compile existing meta-analyses to provide analysis of the multiple postoperative outcomes in a comparison of open-transforaminal lumbar interbody fusions (O-TLIFs) versus minimally invasive transforaminal interbody fusions (MI-TLIFs).

SUMMARY OF BACKGROUND DATA

TLIF is the standard surgical intervention for spinal fusion in degenerative spinal diseases. The comparative effectiveness of MI-TLIFs and O-TLIFs remains controversial.

METHODS

A literature search was conducted in the PubMed, Scopus, and Web of Science databases. Titles and abstracts were initially screened, followed by a full-text review based on the inclusion criteria. Twenty articles were deemed eligible for the umbrella review. Data extraction and quality assessment using A Measurement Tool to Assess Systematic Reviews were performed. Effect sizes of the outcomes of interest from primary studies included in the meta-analyses were repooled. Repooling and stratification of the credibility of the evidence were performed using the R package metaumbrella. The pooled effect sizes were compared and interpreted using equivalent Hedges' g values.

RESULTS

When the meta-analyses were pooled, MI-TLIF was found to have a shorter length of stay, less blood loss, and a higher radiation exposure time, with a highly suggestive level of evidence. Data regarding less postoperative drainage, infections, and Oswestry disability index for MI-TLIF were supported by weak evidence. Conversely, data regarding other postoperative outcomes were nonsignificant to draw any conclusions.

CONCLUSION

Our umbrella review provides a comprehensive overview of the relevant strengths and weaknesses of each surgical technique. This overview revealed that MI-TLIF had better outcomes in terms of length of stay, blood loss, postoperative drainage, infections, and Oswestry disability index when compared with those of O-TLIF. However, O-TLIF had a better outcome for radiation exposure when compared with MI-TLIF.

Prognosis at Your Fingertips: A Machine Learning-Based Web Application for Outcome Prediction in Acute Traumatic Epidural Hematoma

Traumatic brain injury (TBI) affects 69 million people worldwide each year, and acute traumatic epidural hematoma (atEDH) is a frequent and severe consequence of TBI. The aim of the study is to use machine learning (ML) algorithms to predict in-hospital death, non-home discharges, prolonged length of stay (LOS), prolonged length of intensive care unit stay (ICU-LOS), and major complications in patients with atEDH and incorporate the resulting ML models into a user-friendly web application for use in the clinical settings. The American College of Surgeons (ACS) Trauma Quality Program (TQP) database was used to identify patients with atEDH. Four ML algorithms (XGBoost, LightGBM, CatBoost, and Random Forest) were utilized, and the best performing models were incorporated into an open-access web application to predict the outcomes of interest. The study found that the ML algorithms had high area under the receiver operating characteristic curve (AUROC) values in predicting outcomes for patients with atEDH. In particular, the algorithms had an AUROC value range of between 0.874 to 0.956 for in-hospital mortality, 0.776 to 0.798 for non-home discharges, 0.737 to 0.758 for prolonged LOS, 0.712 to 0.774 for prolonged ICU-LOS, and 0.674 to 0.733 for major complications. The following link will take users to the open-access web application designed to generate predictions for individual patients based on their characteristics: huggingface.co/spaces/MSHS-Neurosurgery-Research/TQP-atEDH. This study aimed to improve the prognostication of patients with atEDH using ML algorithms and developed a web application for easy integration in clinical practice. It found that ML algorithms can aid in risk stratification and have significant potential for predicting in-hospital outcomes. Results demonstrated excellent performance for predicting in-hospital death and fair performance for non-home discharges, prolonged LOS and ICU-LOS, and poor performance for major complications.

Cervical laminoplasty versus laminectomy and fusion: An umbrella review of postoperative outcomes

While multiple studies exist comparing cervical laminoplasty (CLP) and posterior cervical laminectomy with fusion (PCF), no clear consensus exists on which intervention is better. An umbrella review helps provide an overall assessment by analyzing a given condition's multiple interventions and outcomes. It integrates all available information on a topic and allows a consensus to be reached on the intervention of choice. A literature search was conducted using specific search criteria in PubMed, Scopus, and Web of Science databases. Titles and abstracts were screened based on inclusion criteria. A full-text review of articles that passed the initial inclusion criteria was performed. Nine meta-analyses were deemed eligible for the umbrella review. Data was extracted on reported variables from these meta-analyses. Subsequent quality assessment using AMSTAR2 and data analysis using the R package metaumbrella were used to determine the significance of postoperative outcomes. When the meta-analyses were pooled, statistically significant differences between CLP and PCF were found for postoperative overall complications rate and postoperative JOA score. PCF was associated with a lower overall complication rate and a higher postoperative JOA score, both supported by a weak level of evidence (class IV). Data regarding all other outcomes were non-significant. Our umbrella review investigates CLP and PCF by providing a comprehensive overview of existing evidence and evaluating inconsistencies within the literature. This umbrella review revealed that PCF had better outcomes for overall complications rate and postoperative JOA than CLP, but they were classified as being of weak significance.

Ultra-early (≤8 hours) surgery for thoracolumbar spinal cord injuries: A systematic review and meta-analysis

Background

The impact of the timing of surgery on neurological recovery in thoracolumbar spinal cord injuries (tSCI) is still a subject of discussion. Accumulating evidence is supporting early decompression (<24 hours) following tSCI. However, the potential advantages of earlier decompression remain uncertain. This systematic review and meta-analysis summarize and analyze the current evidence on the effectiveness of ultra-early decompression surgery on clinical outcomes following tSCI.

Methods

A search was conducted in the electronic databases Medline, Embase, Scopus, and Web of Science from their inception until May 2022 for human studies. Groups were stratified into ultra-early (surgery within 8 hours of injury) vs control group operated >8 hours of injury. The authors included the study data from their institutional case series of thoracolumbar spinal cord injury from 2015 to 2018. An arm-based meta-analysis was performed on all studies using the R Studio. For studies that qualified, a contrast-based meta-analysis was also performed with a standardized mean difference (SMD). Outcomes were reported as effect size, treatment effect, and effect difference, all with 95% confidence intervals (CI).

Results

Of the 133 patients, 74.4% patients were male. 76 (57.1%) underwent decompression ≤8 hours, while 57 (42.9%) underwent decompression >8 hours from injury. Quantitative analysis using the SMD model showed a significant difference in mean AIS improvement in the ultra-early group (Effect size 1.15 [0.62-1.67], p<.0001). On arm-based meta-analysis, a statistically significant treatment effect was found for the ultra-early arm (1.25 [0.91-1.67]), while > 8-hour arm did not show significance (0.30 [-0.08-0.71]). There was a statistically significant effect difference between the two arms (0.96 [0.49-1.48]).

Conclusions

This study observed a significant improvement in the mean AIS score in patients undergoing decompression within 8 hours of tSCI. Given the scant literature regarding ultra-early decompression of tSCI, this study solidifies the need to further explore the role of early interventions for tSCIs to improve patient outcomes.

Precision medicine for traumatic cervical spinal cord injuries: accessible and interpretable machine learning models to predict individualized in-hospital outcomes

BACKGROUND CONTEXT

A traumatic spinal cord injury (SCI) can cause temporary or permanent motor and sensory impairment, leading to serious short and long-term consequences that can result in significant morbidity and mortality. The cervical spine is the most commonly affected area, accounting for about 60% of all traumatic SCI cases.

PURPOSE

This study aims to employ machine learning (ML) algorithms to predict various outcomes, such as in-hospital mortality, nonhome discharges, extended length of stay (LOS), extended length of intensive care unit stay (ICU-LOS), and major complications in patients diagnosed with cervical SCI (cSCI).

STUDY DESIGN

Our study was a retrospective machine learning classification study aiming to predict the outcomes of interest, which were binary categorical variables, in patients diagnosed with cSCI.

PATIENT SAMPLE

The data for this study were obtained from the American College of Surgeons (ACS) Trauma Quality Program (TQP) database, which was queried to identify patients who suffered from cSCI between 2019 and 2021.

OUTCOME MEASURES

The outcomes of interest of our study were in-hospital mortality, nonhome discharges, prolonged LOS, prolonged ICU-LOS, and major complications. The study evaluated the models' performance using both graphical and numerical methods. The receiver operating characteristic (ROC) and precision-recall curves (PRC) were used to assess model performance graphically. Numerical evaluation metrics included AUROC, balanced accuracy, weighted area under PRC (AUPRC), weighted precision, and weighted recall.

METHODS

The study employed data from the American College of Surgeons (ACS) Trauma Quality Program (TQP) database to identify patients with cSCI. Four ML algorithms, namely XGBoost, LightGBM, CatBoost, and Random Forest, were utilized to develop predictive models. The most effective models were then incorporated into a publicly available web application designed to forecast the outcomes of interest.

RESULTS

There were 71,661 patients included in the analysis for the outcome mortality, 67,331 for the outcome nonhome discharges, 76,782 for the outcome prolonged LOS, 26,615 for the outcome prolonged ICU-LOS, and 72,132 for the outcome major complications. The algorithms exhibited an AUROC value range of 0.78 to 0.839 for in-hospital mortality, 0.806 to 0.815 for nonhome discharges, 0.679 to 0.742 for prolonged LOS, 0.666 to 0.682 for prolonged ICU-LOS, and 0.637 to 0.704 for major complications. An open access web application was developed as part of the study, which can generate predictions for individual patients based on their characteristics.

CONCLUSIONS

Our study suggests that ML models can be valuable in assessing risk for patients with cervical cSCI and may have considerable potential for predicting outcomes during hospitalization. ML models demonstrated good predictive ability for in-hospital mortality and nonhome discharges, fair predictive ability for prolonged LOS, but poor predictive ability for prolonged ICU-LOS and major complications. Along with these promising results, the development of a user-friendly web application that facilitates the integration of these models into clinical practice is a significant contribution of this study. The product of this study may have significant implications in clinical settings to personalize care, anticipate outcomes, facilitate shared decision making and informed consent processes for cSCI patients.

The Frequency of Symptoms in Patients With a Diagnosis of Degenerative Cervical Myelopathy: Results of a Scoping Review

STUDY DESIGN

Delayed diagnosis of degenerative cervical myelopathy (DCM) is associated with reduced quality of life and greater disability. Developing diagnostic criteria for DCM has been identified as a top research priority.

OBJECTIVES

This scoping review aims to address the following questions: What is the diagnostic accuracy and frequency of clinical symptoms in patients with DCM?

METHODS

A scoping review was conducted using a database of all primary DCM studies published between 2005 and 2020. Studies were included if they (i) assessed the diagnostic accuracy of a symptom using an appropriate control group or (ii) reported the frequency of a symptom in a cohort of DCM patients.

RESULTS

This review identified three studies that discussed the diagnostic accuracy of various symptoms and included a control group. An additional 58 reported on the frequency of symptoms in a cohort of patients with DCM. The most frequent and sensitive symptoms in DCM include unspecified paresthesias (86%), hand numbness (82%) and hand paresthesias (79%). Neck and/or shoulder pain was present in 51% of patients with DCM, whereas a minority had back (19%) or lower extremity pain (10%). Bladder dysfunction was uncommon (38%) although more frequent than bowel (23%) and sexual impairment (4%). Gait impairment is also commonly seen in patients with DCM (72%).

CONCLUSION

Patients with DCM present with many different symptoms, most commonly sensorimotor impairment of the upper extremities, pain, bladder dysfunction and gait disturbance. If patients present with a combination of these symptoms, further neuroimaging is indicated to confirm the diagnosis of DCM.

Assessment of Computed Tomography Perfusion Research Landscape: A Topic Modeling Study

The number of scholarly articles continues to rise. The continuous increase in scientific output poses a challenge for researchers, who must devote considerable time to collecting and analyzing these results. The topic modeling approach emerges as a novel response to this need. Considering the swift advancements in computed tomography perfusion (CTP), we deem it essential to launch an initiative focused on topic modeling. We conducted a comprehensive search of the Scopus database from 1 January 2000 to 16 August 2023, to identify relevant articles about CTP. Using the BERTopic model, we derived a group of topics along with their respective representative articles. For the 2020s, linear regression models were used to identify and interpret trending topics. From the most to the least prevalent, the topics that were identified include "Tumor Vascularity", "Stroke Assessment", "Myocardial Perfusion", "Intracerebral Hemorrhage", "Imaging Optimization", "Reperfusion Therapy", "Postprocessing", "Carotid Artery Disease", "Seizures", "Hemorrhagic Transformation", "Artificial Intelligence", and "Moyamoya Disease". The model provided insights into the trends of the current decade, highlighting "Postprocessing" and "Artificial Intelligence" as the most trending topics.

Development of personalized machine learning-based prediction models for short-term postoperative outcomes in patients undergoing cervical laminoplasty

PURPOSE

By predicting short-term postoperative outcomes before surgery, patients undergoing cervical laminoplasty (CLP) surgery could benefit from more accurate patient care strategies that could reduce the likelihood of adverse outcomes. With this study, we developed a series of machine learning (ML) models for predicting short-term postoperative outcomes and integrated them into an open-source online application.

METHODS

National surgical quality improvement program database was utilized to identify individuals who have undergone CLP surgery. The investigated outcomes were prolonged length of stay (LOS), non-home discharges, 30-day readmissions, unplanned reoperations, and major complications. ML models were developed and implemented on a website to predict these three outcomes.

RESULTS

A total of 1740 patients that underwent CLP were included in the analysis. Performance evaluation indicated that the top-performing models for each outcome were the models built with TabPFN and LightGBM algorithms. The TabPFN models yielded AUROCs of 0.830, 0.847, and 0.858 in predicting non-home discharges, unplanned reoperations, and major complications, respectively. The LightGBM models yielded AUROCs of 0.812 and 0.817 in predicting prolonged LOS, and 30-day readmissions, respectively.

CONCLUSION

The potential of ML approaches to predict postoperative outcomes following spine surgery is significant. As the volume of data in spine surgery continues to increase, the development of predictive models as clinically relevant decision-making tools could significantly improve risk assessment and prognosis. Here, we present an accessible predictive model for predicting short-term postoperative outcomes following CLP intended to achieve the stated objectives.

The value of Clinical signs in the diagnosis of Degenerative Cervical Myelopathy - A Systematic review and Meta-analysis

STUDY DESIGN

Delayed diagnosis of degenerative cervical myelopathy (DCM) is likely due to a combination of its subtle symptoms, incomplete neurological assessments by clinicians and a lack of public and professional awareness. Diagnostic criteria for DCM will likely facilitate earlier referral for definitive management.

OBJECTIVES

This systematic review aims to determine (i) the diagnostic accuracy of various clinical signs and (ii) the association between clinical signs and disease severity in DCM?

METHODS

A search was performed to identify studies on adult patients that evaluated the diagnostic accuracy of a clinical sign used for diagnosing DCM. Studies were also included if they assessed the association between the presence of a clinical sign and disease severity. The QUADAS-2 tool was used to evaluate the risk of bias of individual studies.

RESULTS

This review identified eleven studies that used a control group to evaluate the diagnostic accuracy of various signs. An additional 61 articles reported on the frequency of clinical signs in a cohort of DCM patients. The most sensitive clinical tests for diagnosing DCM were the Tromner and hyperreflexia, whereas the most specific tests were the Babinski, Tromner, clonus and inverted supinator sign. Five studies evaluated the association between the presence of various clinical signs and disease severity. There was no definite association between Hoffmann sign, Babinski sign or hyperreflexia and disease severity.

CONCLUSION

The presence of clinical signs suggesting spinal cord compression should encourage health care professionals to pursue further investigation, such as neuroimaging to either confirm or refute a diagnosis of DCM.

Prognosis Individualized: Survival predictions for WHO grade II and III gliomas with a machine learning-based web application

WHO grade II and III gliomas demonstrate diverse biological behaviors resulting in variable survival outcomes. In the context of glioma prognosis, machine learning (ML) approaches could facilitate the navigation through the maze of factors influencing survival, aiding clinicians in generating more precise and personalized survival predictions. Here we report the utilization of ML models in predicting survival at 12, 24, 36, and 60 months following grade II and III glioma diagnosis. From the National Cancer Database, we analyze 10,001 WHO grade II and 11,456 grade III cranial gliomas. Using the area under the receiver operating characteristic (AUROC) values, we deploy the top-performing models in a web application for individualized predictions. SHapley Additive exPlanations (SHAP) enhance the interpretability of the models. Top-performing predictive models are the ones built with LightGBM and Random Forest algorithms. For grade II gliomas, the models yield AUROC values ranging from 0.813 to 0.896 for predicting mortality across different timeframes, and for grade III gliomas, the models yield AUROCs ranging from 0.855 to 0.878. ML models provide individualized survival forecasts for grade II and III glioma patients across multiple clinically relevant time points. The user-friendly web application represents a pioneering digital tool to potentially integrate predictive analytics into neuro-oncology clinical practice, to empower prognostication and personalize clinical decision-making.

Advancing personalized prognosis in atypical and anaplastic meningiomas through interpretable machine learning models

PURPOSE

The primary purpose of this study was to utilize machine learning (ML) models to create a web application that can predict survival outcomes for patients diagnosed with atypical and anaplastic meningiomas.

METHODS

In this retrospective cohort study, patients diagnosed with WHO grade II and III meningiomas were selected from the National Cancer Database (NCDB) to analyze survival outcomes at 12, 36, and 60 months. Five machine learning algorithms - TabPFN, TabNet, XGBoost, LightGBM, and Random Forest were employed and optimized using the Optuna library for hyperparameter tuning. The top-performing models were then deployed into our web-based application.

RESULTS

From the NCDB, 12,197 adult patients diagnosed with histologically confirmed WHO grade II and III meningiomas were retrieved. The mean age was 61 (± 20), and 6,847 (56.1%) of these were females. Performance evaluation indicated that the top-performing models for each outcome were the models built with the TabPFN algorithm. The TabPFN models yielded area under the receiver operating characteristic (AUROC) values of 0.805, 0.781, and 0.815 in predicting 12-, 36-, and 60-month mortality, respectively.

CONCLUSION

With the continuous growth of neuro-oncology data, ML algorithms act as key tools in predicting survival outcomes for WHO grade II and III meningioma patients. By incorporating these interpretable models into a web application, we can practically utilize them to improve risk evaluation and prognosis for meningioma patients.

Interpretable machine learning models to predict short-term postoperative outcomes following posterior cervical fusion

By predicting short-term postoperative outcomes before surgery, patients who undergo posterior cervical fusion (PCF) surgery may benefit from more precise patient care plans that reduce the likelihood of unfavorable outcomes. We developed machine learning models for predicting short-term postoperative outcomes and incorporate these models into an open-source web application in this study. The American College of Surgeons National Surgical Quality Improvement Program database was used to identify patients who underwent PCF surgery. Prolonged length of stay, non-home discharges, and readmissions were the three outcomes that were investigated. To predict these three outcomes, machine learning models were developed and incorporated into an open access web application. A total of 6277 patients that underwent PCF surgery were included in the analysis. The most accurately predicted outcome in terms of the area under the receiver operating characteristic curve (AUROC) was the non-home discharges with a mean AUROC of 0.812, and the most accurately predicting algorithm in terms of AUROC was the LightGBM algorithm with a mean AUROC of 0.766. The following URL will take users to the open access web application created to provide predictions for individual patients based on their characteristics: https://huggingface.co/spaces/MSHS-Neurosurgery-Research/NSQIP-PCF. Machine learning techniques have a significant potential for predicting postoperative outcomes following PCF surgery. The development of predictive models as clinically useful decision-making tools may significantly improve risk assessment and prognosis as the amount of data in spinal surgery keeps growing. Here, we present predictive models for PCF surgery that are meant to accomplish the aforementioned goals and make them publicly available.

Evaluating Adult Idiopathic Scoliosis as an Independent Risk Factor for Critical Illness in SARS-CoV-2 Infection

BACKGROUND

Thoracic spinal deformities may reduce chest wall compliance, leading to respiratory complications. The first SARS-CoV-2 (L-variant) strain caused critical respiratory illness, especially in vulnerable patients. This study investigates the association between scoliosis and SARS-CoV-2 (COVID-19) disease course severity.

METHODS

Clinical data of 129 patients treated between March 2020 to June 2021 who received a positive COVID-19 polymerase chain reaction result from Mount Sinai and had a scoliosis ICD-10 code (M41.0-M41.9) was retrospectively analyzed. Degree of coronal plane scoliosis on imaging was confirmed by 2 independent measurers and grouped into no scoliosis (Cobb angle <10°), mild (10°-24°), moderate (25°-39°), and severe (>40°) cohorts. Baseline characteristics were compared, and a multivariable logistic regression controlling for clinically significant comorbidities examined the significance of scoliosis as an independent risk factor for hospitalization, intensive care unit (ICU) admission, acute respiratory distress syndrome (ARDS), mechanical ventilation, and mortality.

RESULTS

The no (n = 42), mild (n = 14), moderate (n = 44), and severe scoliosis (n = 29) cohorts differed significantly only in age (P = 0.026). The percentage of patients hospitalized (P = 0.59), admitted to the ICU (P = 0.33), developing ARDS (P = 0.77), requiring mechanical ventilation (P = 1.0), or who expired (P = 0.77) did not significantly differ between cohorts. The scoliosis cohorts did not have a significantly higher likelihood of hospital admission (mild P = 0.19, moderate P = 0.67, severe P = 0.98), ICU admission (P = 0.97, P = 0.94, P = 0.22), ARDS (P = 0.87, P = 0.74, P = 0.94), mechanical ventilation (P = 0.73, P = 0.69, P = 0.70), or mortality (P = 0.74, P = 0.87, P = 0.66) than the no scoliosis cohort.

CONCLUSIONS

Scoliosis was not an independent risk factor for critical COVID-19 illness. No trends indicated any consistent effect of degree of scoliosis on increased adverse outcome likelihood.

Machine learning-based prediction of short-term adverse postoperative outcomes in cervical disc arthroplasty patients

OBJECTIVE

This study aimed to assess the effectiveness of machine learning (ML) algorithms in predicting short-term adverse postoperative outcomes after cervical disc arthroplasty (CDA) and to create a user-friendly and accessible tool for this purpose.

METHODS

The American College of Surgeons National Surgical Quality Improvement Program database was used to identify patients who underwent CDA. The outcome of interest was the combined occurrence of adverse events in the short-term postoperative period, including prolonged stay, major complications, non-home discharges, and 30-day readmissions. To predict the combined outcome of interest, short-term adverse postoperative outcomes, four different ML algorithms were utilized to develop predictive models, and these models were incorporated into an open access web application.

RESULTS

A total of 6604 patients that underwent CDA were included in the analysis. The mean area under the receiver operating characteristic curve (AUROC) and accuracy were 0.814 and 87.8% for all algorithms. SHapley Additive exPlanations (SHAP) analyses revealed that white race was the most important predictor variable for all four algorithms. The following URL will take users to the open access web application created to provide predictions for individual patients based on their characteristics: huggingface.co/spaces/MSHS-Neurosurgery-Research/NSQIP-CDA.

CONCLUSIONS

ML approaches have the potential to predict postoperative outcomes after CDA surgery. As the amount of data in spinal surgery grows, the development of predictive models as clinically useful decision-making tools may significantly improve risk assessment and prognosis. We present and make publicly available predictive models for CDA intended to achieve the goals mentioned above.

The Advent of Generative Language Models in Medical Education

Artificial intelligence (AI) and generative language models (GLMs) present significant opportunities for enhancing medical education, including the provision of realistic simulations, digital patients, personalized feedback, evaluation methods, and the elimination of language barriers. These advanced technologies can facilitate immersive learning environments and enhance medical students' educational outcomes. However, ensuring content quality, addressing biases, and managing ethical and legal concerns present obstacles. To mitigate these challenges, it is necessary to evaluate the accuracy and relevance of AI-generated content, address potential biases, and develop guidelines and policies governing the use of AI-generated content in medical education. Collaboration among educators, researchers, and practitioners is essential for developing best practices, guidelines, and transparent AI models that encourage the ethical and responsible use of GLMs and AI in medical education. By sharing information about the data used for training, obstacles encountered, and evaluation methods, developers can increase their credibility and trustworthiness within the medical community. In order to realize the full potential of AI and GLMs in medical education while mitigating potential risks and obstacles, ongoing research and interdisciplinary collaboration are necessary. By collaborating, medical professionals can ensure that these technologies are effectively and responsibly integrated, contributing to enhanced learning experiences and patient care.

Comparison of Frailty Indices and the Charlson Comorbidity Index in Traumatic Brain Injury

BACKGROUND

Comorbidity scales for outcome prediction in traumatic brain injury (TBI) include the 5-component modified Frailty Index (mFI-5), the 11-component modified Frailty Index (mFI-11), and the Charlson Comorbidity Index (CCI).

OBJECTIVE

To compare the accuracy in predicting clinical outcomes in TBI of mFI-5, mFI-11, and CCI.

METHODS

The National Trauma Data Bank (NTDB) of the American College of Surgeons (ACS) was utilized to study patients with isolated TBI for the years of 2017 and 2018. After controlling for age and injury severity, individual multivariable logistic regressions were conducted with each of the 3 scales (mFI-5, mFI-11, and CCI) against predefined outcomes, including any complication, home discharge, facility discharge, and mortality.

RESULTS

All 3 scales demonstrated adequate internal consistency throughout their individual components (0.63 for mFI-5, 0.60 for CCI, and 0.56 for mFI-11). Almost all studied complications were significantly more likely in frail patients. mFI-5 and mFI-11 had similar areas under the curve (AUC) for all outcomes, while CCI had lower AUCs (0.62-0.61-0.53 for any complication, 0.72-0.72-0.52 for home discharge, 0.78-0.78-0.53 for facility discharge, and 0.71-0.70-0.52 for mortality, respectively).

CONCLUSION

mFI-5 and mFI-11 demonstrated similar accuracy in predicting any complication, home discharge, facility discharge, and mortality in TBI patients across the NTDB. In addition, CCI's performance was poor for the aforementioned metrics. Since mFI-5 is simpler, yet as accurate as the 2 other scales, it may be the most practical both for clinical practice and for future studies with the NTDB.

Embracing Large Language Models for Medical Applications: Opportunities and Challenges

Large language models (LLMs) have the potential to revolutionize the field of medicine by, among other applications, improving diagnostic accuracy and supporting clinical decision-making. However, the successful integration of LLMs in medicine requires addressing challenges and considerations specific to the medical domain. This viewpoint article provides a comprehensive overview of key aspects for the successful implementation of LLMs in medicine, including transfer learning, domain-specific fine-tuning, domain adaptation, reinforcement learning with expert input, dynamic training, interdisciplinary collaboration, education and training, evaluation metrics, clinical validation, ethical considerations, data privacy, and regulatory frameworks. By adopting a multifaceted approach and fostering interdisciplinary collaboration, LLMs can be developed, validated, and integrated into medical practice responsibly, effectively, and ethically, addressing the needs of various medical disciplines and diverse patient populations. Ultimately, this approach will ensure that LLMs enhance patient care and improve overall health outcomes for all.

A Machine Learning-Based Online Prediction Tool for Predicting Short-Term Postoperative Outcomes Following Spinal Tumor Resections

Background: Preoperative prediction of short-term postoperative outcomes in spinal tumor patients can lead to more precise patient care plans that reduce the likelihood of negative outcomes. With this study, we aimed to develop machine learning algorithms for predicting short-term postoperative outcomes and implement these models in an open-source web application. Methods: Patients who underwent surgical resection of spinal tumors were identified using the American College of Surgeons, National Surgical Quality Improvement Program. Three outcomes were predicted: prolonged length of stay (LOS), nonhome discharges, and major complications. Four machine learning algorithms were developed and integrated into an open access web application to predict these outcomes. Results: A total of 3073 patients that underwent spinal tumor resection were included in the analysis. The most accurately predicted outcomes in terms of the area under the receiver operating characteristic curve (AUROC) was the prolonged LOS with a mean AUROC of 0.745 The most accurately predicting algorithm in terms of AUROC was random forest, with a mean AUROC of 0.743. An open access web application was developed for getting predictions for individual patients based on their characteristics and this web application can be accessed here: huggingface.co/spaces/MSHS-Neurosurgery-Research/NSQIP-ST. Conclusion: Machine learning approaches carry significant potential for the purpose of predicting postoperative outcomes following spinal tumor resections. Development of predictive models as clinically useful decision-making tools may considerably enhance risk assessment and prognosis as the amount of data in spinal tumor surgery continues to rise.

Targeting earlier diagnosis: What symptoms come first in Degenerative Cervical Myelopathy?

BACKGROUND

Degenerative cervical myelopathy (DCM) is a common and disabling condition. Early effective treatment is limited by late diagnosis. Conventional descriptions of DCM focus on motor and sensory limb disability, however, recent work suggests the true impact is much broader. This study aimed to characterise the symptomatic presentation of DCM from the perspective of people with DCM and determine whether any of the reported symptoms, or groups of symptoms, were associated with early diagnosis.

METHODS

An internet survey was developed, using an established list of patient-reported effects. Participants (N = 171) were recruited from an online community of people with DCM. Respondents selected their current symptoms and primary presenting symptom. The relationship of symptoms and their relationship to time to diagnosis were explored. This included symptoms not commonly measured today, termed 'non-conventional' symptoms.

RESULTS

All listed symptoms were experienced by >10% of respondents, with poor balance being the most commonly reported (84.2%). Non-conventional symptoms accounted for 39.7% of symptomatic burden. 55.4% of the symptoms were reported as an initial symptom, with neck pain the most common (13.5%). Non-conventional symptoms accounted for 11.1% of initial symptoms. 79.5% of the respondents were diagnosed late (>6 months). Heavy legs was the only initial symptom associated with early diagnosis.

CONCLUSIONS

A comprehensive description of the self-reported effects of DCM has been established, including the prevalence of symptoms at disease presentation. The experience of DCM is broader than suggested by conventional descriptions and further exploration of non-conventional symptoms may support earlier diagnosis.

International Survey of Antiseizure Medication Use in Patients with Complicated Mild Traumatic Brain Injury: A New York Neurotrauma Consortium Study

BACKGROUND

Seizures and epilepsy after traumatic brain injury (TBI) negatively affect quality of life and longevity. Antiseizure medication (ASM) prophylaxis after severe TBI is associated with improved outcomes; these medications are rarely used in mild TBI. However, a paucity of research is available to inform ASM use in complicated mild TBI (cmTBI) and no empirically based clinical care guidelines for ASM use in cmTBI exist. We aim to identify seizure prevention and management strategies used by clinicians experienced in treating patients with cmTBI to characterize standard care and inform a systematic approach to clinical decision making regarding ASM prophylaxis.

METHODS

We recruited a multidisciplinary international cohort through professional organizational listservs and social media platforms. Our questionnaire assessed factors influencing ASM prophylaxis after cmTBI at the individual, institutional, and health system-wide levels.

RESULTS

Ninety-two providers with experience managing cmTBI completed the survey. We found a striking diversity of ASM use in cmTBI, with 30% of respondents reporting no/infrequent use and 42% reporting frequent use; these tendencies did not differ by provider or institutional characteristics. Certain conditions universally increased or decreased the likelihood of ASM use and represent consensus. Based on survey results, ASMs are commonly used in patients with cmTBI who experience acute secondary seizure or select positive neuroimaging findings; we advise caution in elderly patients and those with concomitant neuropsychiatric illness.

CONCLUSIONS

This study is the first to characterize factors influencing clinical decision making in ASM prophylaxis after cmTBI based on multidisciplinary multicenter provider practices. Prospective controlled studies are necessary to inform standardized guideline development.

Racial and Socioeconomic Disparities in Neurotrauma: Research Priorities in the New York Metropolitan Area through a Global Neurosurgery Paradigm

The New York Neurotrauma Consortium (NYNC) is a nascent multidisciplinary research and advocacy organization based in the New York Metropolitan Area (NYMA). It aims to advance health equity and optimize outcomes for traumatic brain and spine injury patients. Given the extensive racial, ethnic, and socioeconomic diversity of the NYMA, global health frameworks aimed at eliminating disparities in neurotrauma may provide a relevant and useful model for the informing research agendas of consortia like the NYNC. In this review, we present a comparative analysis of key health disparities in traumatic brain injury (TBI) that persist in the NYMA as well as in low- and middle-income countries (LMIC). Examples include: (a) inequitable access to quality care due to fragmentation of healthcare systems, (b) barriers to effective prehospital care for TBI, and (c) socioeconomic challenges faced by patients and their families during the subacute and chronic post-injury phases of TBI care. This review presents strategies to address each area of health disparity based on previous studies conducted in both LMIC and high-income country (HIC) settings. Increased awareness of healthcare disparities, education of healthcare professionals, effective policy advocacy for systemic changes, and fostering racial diversity of the trauma care workforce can guide the development of trauma care systems in the NYMA that are free of racial and related healthcare disparities.

Delayed Cranial Decompression Rates After Initiation of Unfractionated Heparin versus Low-Molecular-Weight Heparin in Traumatic Brain Injury

BACKGROUND

Both unfractionated heparin (UH) and low-molecular-weight heparin (LMWH) are routinely used prophylactically after traumatic brain injury (TBI) to prevent deep vein thrombosis (DVT). Their comparative risk for development or worsening of intracranial hemorrhage necessitating cranial decompression is unclear. Furthermore, the absence of a specific antidote for LMWH may lead to UH being used more often for high-risk patients. This study aims to compare the incidence of delayed cranial decompression occurring after initiation of prophylactic UH versus LMWH using the National Trauma Data Bank.

METHODS

Cranial decompression procedures included craniotomy and craniectomy. Multiple imputation was used for missing data. Propensity score matching was used to account for selection bias between UH and LMWH. The 1:1 matched groups were compared using logistic regression for the primary outcome of postprophylaxis cranial decompression.

RESULTS

A total of 218,594 patients with TBI were included, with 61,998 (28.3%) receiving UH and 156,596 (71.7%) receiving LMWH as DVT prophylaxis. The UH group had higher patient age, body mass index, comorbidity rates, Injury Severity Score, and worse motor Glasgow Coma Scale score. After the UH and LMWH groups were matched for these factors, logistic regression showed lower rates of postprophylaxis cranial decompression for the LMWH group (odds ratio, 0.13; 95% confidence interval, 0.11-0.16; P < 0.001).

CONCLUSIONS

Despite the absence of a specific antidote, LMWH was associated with lower rates of need for post-DVT-prophylaxis in craniotomy/craniectomy. This finding questions the notion of UH being safer for patients with TBI because it can be readily reversed. Randomized studies are needed to elucidate causality.

Prognostic Factors for Stage 3 Acute Kidney Injury in Isolated Serious Traumatic Brain Injury

BACKGROUND

Stage 3 acute kidney injury (AKI) has been observed to develop following serious traumatic brain injury (TBI) and is associated with worse outcomes, though its incidence is not consistently established. This study aims to report the incidence of stage 3 AKI in serious isolated TBI in a large, national trauma database, and explore associated predictive factors.

METHODS

This was a retrospective cohort study using 2015-2018 data from the American College of Surgeons Trauma Quality Improvement Program (ACS-TQIP), a national database of trauma patients. Adult trauma patients admitted to the hospital with isolated serious TBI were included. Variables relating to demographics, comorbidities, vitals, hospital presentation, and course of stay were assessed. Imputed multivariable logistic regression assessed factors predictive of stage 3 AKI development.

RESULTS

A total of 342,675 patients with isolated serious TBI were included, 1,585 (0.5%) of whom developed stage 3 AKI. Variables associated with stage 3 AKI in multivariable analysis were older age, male sex, Black race, higher BMI, history of hypertension, diabetes, peripheral artery disease, chronic kidney disease, higher injury severity score, higher heart rate on arrival, lower oxygen saturation and motor Glasgow coma scale (GCS), admission to the intensive care unit (ICU) or operating room, development of catheter-associated urinary tract infections (CAUTI) or acute respiratory distress syndrome (ARDS), longer ICU stay and ventilation duration.

CONCLUSIONS

Stage 3 AKI occurred in 0.5% of serious TBI cases. Complications of ARDS and CAUTI are more likely to co-occur with stage 3 AKI in serious TBI patients.

Minimally invasive endoscopic approach to perform complete coccygectomy in patients with chronic refractory coccydynia: illustrative case

BACKGROUND

Coccydynia refers to debilitating pain in the coccygeal region of the spine. Treatment strategies range from conservative measures (e.g., ergonomic adaptations, physical therapy, nerve block injections) to partial or complete removal of the coccyx (coccygectomy). Because the surgical intervention is situated in a high-pressure location close to the anus, a possible complication is the formation of sacral pressure ulcers and infection at the incision site.

OBSERVATIONS

In this case report, the authors presented a minimally invasive, fully endoscopic approach to safely perform complete coccygectomy for treatment of refractory posttraumatic coccydynia.

LESSONS

Although this is a single case report, the authors hope that this novel endoscopic approach may achieve improved wound healing, reduced infection rates, and lower risk of penetration injury to retroperitoneal organs in patients requiring coccygectomy.

Availability and Readability of Spinal Cord Injury Online Information Materials for Spanish Speaking Population in Neurosurgical Academic Programs: A Nationwide Study

BACKGROUND

Recent publications have demonstrated that information has been transmitted inappropriately to the lay person in different pathologies. This limitation is also observed in Spanish language. We evaluate the availability and readability of online patient education material (PEM) on spinal cord injury (SCI) information for the Spanish-speaking population from academic neurosurgery residency programs in the United States.

METHODS

This is a descriptive analysis of online SCI PEM from neurosurgical residency programs websites. We assess the availability of information in Spanish using a modification of a previously published classification. To assess accessibility, we calculated the time spent and the number of clicks to find the information in Spanish. We calculated the readability of the material using the "Indice Flesch-Szigriszt" (INFLESZ), which determines the difficulty of readability of health-related material in Spanish.

RESULTS

A total of 116 accredited neurosurgery residency programs comprised our cohort. Ten (9%) programs had available "mirrored" information in Spanish from its original version in English, 9 (8.1%) used a translation software, 79 (71.2%) provide interpreter services, and 3 (2%) did not have written information or information about translation services. A mean of 72.9 seconds (SD +/- 71.2) were required to have access to the Spanish information or contact information for translation services. Twelve (57.1%) websites with written Spanish information had an INFLESZ score above 55.00, which translates as an appropriate readability level for the general population.

CONCLUSIONS

More than half of the academic neurosurgery programs or affiliated hospital websites do not provide written informative material about SCI in Spanish. When available, the information is not always transmitted with a level of readability appropriate for the layperson. Most of the websites provide translation or interpreter services that are not directly related to SCI.

Dome Laminotomies at Adjacent Segments in Cervical Laminoplasty

BACKGROUND

Cervical laminoplasty is an established and effective surgical treatment for neurologic dysfunction associated with cervical myelopathy. "Dome laminotomies" involve undercutting the laminae adjacent to the laminoplasty levels to decompress and prevent spinal cord kinking on the lamina edges. The technique allows for a decrease in the number of instrumented laminae, smaller surgical exposure, and preservation of muscular attachments at the top of C2 and C7. We investigated whether dome laminotomies are associated with satisfactory neurologic and pain outcome.

METHODS

This study involved a retrospective review of consecutive patients treated at a single institution between November 2015 and September 2018. The patients underwent a C3-C6 laminoplasty with dome laminotomies of the caudal edge of C2 and the cranial edge of C7 lamina. Postoperative evaluations of pain, myelopathy symptoms, and complications occurred at early (mean, ∼2 months) and late (mean, ∼15 months) time points.

RESULTS

Twenty-one patients were enrolled (mean age, 62 years). Mean axial pain score improved significantly at both the early (P = .02) and late (P = .045) postoperative evaluations compared with the mean baseline pain score. A total of 92% of patients experienced resolution of baseline hand dysfunction at the early postoperative follow-up, and 84% maintained it at the late follow-up. Two-thirds of patients experienced (late) significant improvement (P < 0.05) in baseline balance impairment. Postoperative response rates for urinary dysfunction were 58% (early) and 42% (late). There were no wound complications, late neurologic deterioration, kyphosis, or C5 palsy.

CONCLUSIONS

C3-C6 laminoplasty with C2 and C7 dome laminotomies was safe, well tolerated, and associated with satisfactory early and late improved neurologic function and decreased pain.