CWCorey T. Walker

Corey T. Walker

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Corey T. Walker has a total of 78 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    ASAREE: An Analytical Sandbox for Agentic AI Research, Engineering, and Experimentation

    by Jay Moran, Philip J. Freda, Attri Ghosh, Corey T. Walker, Miguel E. Hernandez, Jason H. Moore

    Summary: Agentic AI platforms enable the engineering of autonomous workflows but are not designed for experimentation and hypothesis testing. ASAREE (Analytical Sandbox for Agentic AI Research, Engineering, and Experimentation), is an open-source platform to address this gap. ASAREE creates agents, connects to MCP servers and tools, and designs factorial experiments through a visual interface or Python SDK. It records a full provenance trace for every run and routes all model calls through a provider-agnostic bridge that supports local deployments, ensuring data privacy. As a use-case, we use ASAREE to evaluate key design choices in a mutli-agent machine learning pipeline. Across a 2 × 2 × 2 factorial design, more advanced models, greater reasoning effort, and critic agent use significantly increased compute time, token use, cost, and feature count without improving predictive performance. The lowest-cost baseline, Claude Sonnet 5 with medium effort and no critic, achieved the highest mean PR AUC while Claude Opus 5 with extra high effort and a critic agent cost 15.5 × more (USD) and ran 13.1 × longer while performing worse on average. These findings highlight ASAREE as a robust framework for evaluating agentic system performance and resource efficiency. Availability and implementation: ASAREE is available on GitHub at: https://github.com/EpistasisLab/ASAREE . Contact: jason.moore@csmc.edu Supplementary Information: Supplementary information is available at https://github.com/EpistasisLab/ASAREE/tree/main/publications/bioinformatics

  • Contemporary Concepts in Cervical Disk Arthroplasty: Implant Design, Constraint, Complications, and Technical Pearls

    by Austin Q. Nguyen, Andrei Demkov, Don Le, Rohan Vemu, Mark L. Prasarn, Corey T. Walker, Hyun W. Bae, Todd Albert, Andrew A. Sama, Sheeraz A. Qureshi, et al.

    » Cervical disk arthroplasty (CDA) is a Food and Drug Administration (FDA)–approved motion-preserving alternative to anterior cervical diskectomy and fusion for 1-level and 2-level degenerative disk diseases, supported by high-level evidence and associated with reduced adjacent segment degeneration and comparable or improved clinical outcomes. » FDA-approved CDA implants differ in biomechanical design, degrees of freedom (DoF), and constraint, affecting motion preservation, segmental control, and long-term kinematics. Devices range from constrained (3 DoF) to semiconstrained and viscoelastic designs (4-6 DoF). Implant articulation and materials impact wear, imaging compatibility, and biomechanical behavior and should be considered during device selection. » Constraint-based classification provides a practical, biomechanics-informed framework for CDA implant selection. Constraint level should be matched to patient pathology. Constrained implants may be preferable in cases of segmental instability, compromised endplate integrity, or poor bone quality. Semiconstrained or viscoelastic devices may offer better outcomes for patients with preserved motion, healthy facets, and high functional demands. » A structured understanding of failure modes is critical for safe and effective CDA use. Complications and failure can result from surgical technique errors (e.g., endplate violation, malposition), biologic response (e.g., osteolysis, heterotopic ossification), or mechanical complications (e.g., implant fracture, subsidence). Technical considerations aid the surgeon in reducing risk of complications.


  • Open Access

    Leveraging automated machine learning to benchmark, deconstruct, and compare frailty indices for predicting adverse spinal surgery outcomes

    by Attri Ghosh, Philip J. Freda, Shane Shahrestani, Alena Orlenko, Justin K. Scheer, Tayo Obafemi-Ajayi, Jason H. Moore, Corey T. Walker

    Frailty indices, composite summary scores of diagnostic codes, labs, or vitals, are widely used to assess patient vulnerability. However, their value in predictive modeling remains uncertain. In a cohort of 2,912 elective lumbar spinal fusions, we retrospectively evaluated two established indices, the Hospital Frailty Risk Score (HFRS) and the electronic Frailty Index (eFI), against a curated set of 63 clinical features from electronic medical records. Using automated machine learning optimized logistic regression models, we compared predictive performances and feature importances of frailty indices, and their individual components, against curated clinical features. The curated clinical feature set consistently outperformed (measured by AUC) both frailty indices and their respective components across three outcomes: discharge disposition, length of stay, and 90-day readmission. While individual frailty components improved performance over their respective summary scores, they underperformed compared to clinical variables. However, integrating select frailty-based components with curated clinical features modestly improved predictive performance. Interaction analysis uncovered nonlinear relationships, such as between hemoglobin and smoking, that frailty indices alone did not capture. These findings highlight that while frailty indices aid broad risk stratification, their composite nature limits utility for precise, patient-specific predictive modeling, supporting hybrid strategies centered on curated variables enhanced with selected frailty elements.


  • Open Access

    CT-Based Characterization of Fracture Patterns in Pediatric Lumbar Spondylolysis

    by Vivien Chan, Robert G. Watkins, Kenneth D. Illingworth, Corey T. Walker, David L. Skaggs

    Background: Spondylolysis is a stress fracture of the pars interarticularis. Precise fracture morphology may help dictate management and dictate whether direct repair with an intralaminar screw is possible. The aim of this study was to characterize the specific fracture patterns of the pars interarticularis in pediatric patients with spondylolysis. Methods: This was a single-center retrospective cohort study. Patients were included if they were aged younger than 21 years with lumbar spondylolysis or Grade 1 isthmic spondylolisthesis on computed tomography (CT) imaging. On sagittal CT, total pars length (mm), inferior edge of inferior articular process to spondylolysis (mm), superior edge to spondylolysis (mm), size of spondylolysis gap (mm), and width of pars (mm) at fracture site were measured. The angle of the pars fractures were characterized in reference to the long axis of the pars. Results: There were 32 patients with 59 total spondylolyses included in this study. The mean age was 15.1 ± 1.9 years, and 44% (n = 14) were female. 7 (22%) patients had grade 1 spondylolisthesis. There were 15 fractures (25%) that had less than 0.5-mm gap, and the remaining 44 spondylolysis had a mean gap size of 2.4 ± 1.3 mm (range 0.9-6.4 mm). The mean total pars length was 37.9 ± 3.7 mm (range 31.6-45.8 mm). The mean measurement from the superior edge to the spondylolysis site was 12.1 ± 2.9 mm (range 6.1-18.8 mm). The average percentage of superior edge to spondylolysis/total pars length was 32.0% (range 17%-52%). Pars fracture angles ranged from 77° to 165° to the long axis of the pars. A majority (n = 40, 68%) of the pars fractures were between 110 and 140°. There were 12 fractures (20%) that were 140°. Conclusions: Pars fractures typically occur approximately one-third of the distance from the superior edge of the pars. However, considerable variability exists in their exact location along the pars. A clearer understanding of these fracture patterns may help refine surgical techniques and improve surgical outcomes for this patient population. Level of Evidence: Level III . See Instructions for Authors for a complete description of levels of evidence.

  • Staged Management of Chronic Atlantoaxial Rotatory Subluxation With Halo-Gravity Traction and Cervical Fusion

    by Adeesya Gausper, Vivien Chan, Kenneth Illingworth, Corey Walker

    Case: A 10-year-old girl presented with a 16-month history of persistent head tilt, restricted neck motion, and neck pain. Imaging demonstrated chronic atlantoaxial rotatory subluxation (AARS) with underlying congenital C1 arch diastasis. Preoperative halo-gravity traction (HGT) was performed over 2.5 weeks, achieving partial realignment, followed by C1 to C2 posterior spinal fusion with iliac crest autograft. Postoperatively, the patient demonstrated stable alignment and resolution of neck pain. Conclusion: Chronic AARS may be successfully treated using a staged approach with preoperative HGT followed by cervical arthrodesis.


  • Open Access

    Magnetic Resonance Imaging-Based Assessment of Bone Quality Using Vertebral Bone Quality (VBQ) Scores in Spine Surgery—A Critical Assessment and Narrative Review

    by Adeesya Gausper, Wende N. Gibbs, Benjamin D. Elder, Justin K. Scheer, Tiffany G. Perry, Suhas K. Etigunta, Andy M. Liu, Alexander Tuchman, Corey T. Walker

    Bone health is a key determinant of success in spine surgery, making preoperative assessment of bone quality essential to optimal surgical risk stratification. Magnetic resonance imaging (MRI)-based vertebral bone quality (VBQ) score offers a novel approach to assess bone health in spine surgery candidates. The ability of MRI to assess bone quality without exposure to ionizing radiation makes it a potentially advantageous alternative to other traditional measures of bone density. VBQ has additionally shown potential to predict adverse outcomes, such as fragility fractures, instrumentation failure, subsidence and proximal junctional kyphosis. Variations of VBQ, such as endplate bone quality, S1 VBQ, and cervical VBQ, provide targeted insights at specific anatomical regions and potentially enhance the predictive accuracy of VBQ. However, clinical application of VBQ is limited by variability in MRI systems, patient-specific factors, and lack of standardized threshold values. This review aims to critically evaluate VBQ scores as an opportunistic, MRI-based assessment of bone health and its potential role in predicting surgical outcomes. While VBQ may provide some valuable insights into bone health, its role in preoperative risk assessment likely remains supplementary and requires further research to establish clinical validity and optimal cutoffs.


  • Reoperation Rates After Lumbar Discectomy in Pediatric Patients

    by Andy M. Bs Liu, Adeesya Ba Gausper, Suhas Bs Etigunta, Karim Shafi, Kenneth Illingworth, David Skaggs, Alexander Tuchman, Corey Walker

    Background: Lumbar disc herniation (LDH) is uncommon in the pediatric population but can cause significant low-back or radicular pain and, at times, neurological deficits. We aimed to study discectomy operations in pediatric patients to provide insight into surgical outcomes that may inform clinical decision-making and patient counseling. Methods: A national insurance claims database (PearlDiver) was queried to identify pediatric patients (<21 y old) who underwent discectomy. Procedures were characterized by demographics features, including age of patient, year, and location. Reoperations was defined as discectomy, re-exploration discectomy, fusion, or laminectomy occurring within 5 years of the initial discectomy. A subsequent parallel analysis looked at reoperations following re-exploration discectomies. Kaplan-Meier survival and cox proportional regression analyzed factors impacting survival postprimary discectomy. Results: A total of 4410 primary discectomy patients were identified, with an overall 12% reoperation rate within 5 years of the initial discectomy. Specifically, patients under 18 and between the ages of 18 and 21 having a reoperation rate of 7% (115 patients) and 15% (406 patients), respectively. The incidence of discectomy operations increased with patient age. 78% of pediatric cases were performed on an outpatient basis compared with 75% of adult cases. The rate of reoperation was highest within the first year postdiscectomy, with nearly half of reoperations occurring in this time. Specifically, reoperation rates were 2.3% at 3 months, 4% at 6 months, 6% at 1 year, 7% at 2 years, and 12% at 5 years. The most common reoperation procedure was another discectomy. The reoperation rates following these revision operations were 2% at 3 months, 4% at 6 months, 6% at 1 year, 10% at 2 years, and 14% by 5 years. Fusion was the most common procedure following a failed revision discectomy (42% at 5 y). The Kaplan-Meier survival analysis similarly showed most procedures occurred in the first 3 years, with obesity and Elixhauser Comorbidity Index inversely correlated with survival. Conclusions: Overall, pediatric reoperation rates following discectomy are 11.8%. Obesity and Elixhauser Comorbidity Index significantly increased risk of reoperation. This study provides real-world, large-scale data that may guide surgeons caring for pediatric patients undergoing microdiscectomy. Level of Evidence: Level III.


  • Preoperative anemia is an unsuspecting driver of machine learning prediction of adverse outcomes after lumbar spinal fusion

    by Attri Ghosh, Philip J. Freda, Shane Shahrestani, Andre E. Boyke, Alena Orlenko, Hyunjun Choi, Nicholas Matsumoto, Tayo Obafemi-Ajayi, Jason H. Moore, Corey T. Walker


  • Lateral Corpectomy for Tumor at L1: A Surgical Technique: 2-Dimensional Operative Video

    by Robin M. Babadjouni, Nakul Narendran, Paal K. Nilssen, Alexander Tuchman, Corey T. Walker

    D woman with a 7-year history of breast cancer who presented with low back pain after running a marathon. To date, the patient had deferred medical management. Imaging revealed Stage IV differentiated invasive ductal carcinoma with extensive bony metastatic disease in multiple areas of the spine. The patient underwent a right-sided minimally invasive retrodiaphragmatic approach to the thoracolumbar junction for L1 corpectomy, placement of an expandable cage, and posterior percutaneous pedicle screw segmental fixation from T11 to L3 with robotic guidance. Patient consent was obtained for the following procedure, and IRB approval was not required for publication of this single patient case report....