Study Design: Retrospective Cohort Study. Objectives: To compare 1-year postoperative outcomes and complication rates between single-level lumbar laminectomy and hemilaminectomy using a large, multicenter, propensity-matched dataset. Methods: We queried the TriNetX global health research network (≥160 million patients) for adults undergoing single-level lumbar decompression between January 2005 and July 2025. Cohorts were defined by CPT codes: laminectomy and hemilaminectomy, with qualifying diagnoses of lumbar disc herniation, spinal stenosis, spondylolisthesis, or radiculopathy. Patients with fusion, prior lumbar surgery, or non-degenerative pathology were excluded. Outcomes included new postoperative events within 1 year: mortality, weakness, pain, sensory loss, cauda equina syndrome, radiculopathy, foot drop, CSF leak, and surgical-site infection. Propensity-score matching balanced demographics and comorbidities. Cox proportional hazards models, Kaplan–Meier curves, and relative risks were calculated. Results: Of 167,177 patients, 80,440 underwent laminectomy and 86,737 hemilaminectomy. After matching, 50,853 patients per cohort were analyzed. One-year mortality was similar (0.57% vs 0.49%, HR 1.20; 95% CI 1.01-1.42; P = 0.045). Laminectomy conferred significantly higher risks of CSF leak (1.41% vs 1.00%; RR 1.41), surgical-site infection (1.45% vs 1.00%; RR 1.45), cauda equina syndrome (0.36% vs 0.22%; RR 1.62), and persistent weakness (4.12% vs 3.67%; RR 1.12). Persistent radiculopathy was modestly less frequent after laminectomy (10.5% vs 12.0%; RR 0.87). Other outcomes, including pain and foot drop, were comparable. Conclusions: Hemilaminectomy was associated with lower perioperative complication rates compared to laminectomy, while laminectomy provided a modest reduction in persistent radiculopathy. These findings highlight a tradeoff between safety and decompressive efficacy, emphasizing the importance of patient-specific surgical selection.

Background: Traumatic brain injury (TBI) imposes a substantial public health burden through long-term physical, cognitive, and psychiatric effects. This includes substance use disorders (SUDs) for which TBI is a demonstrated risk factor; however, prior studies have not comprehensively compared relative incidences of SUD subtypes post-TBI or differences between intracranial hemorrhage (ICH) and non-ICH TBI in patients without prior SUD history. This global retrospective analysis using the TriNetX database aims to quantify new-onset SUD incidence post-TBI in the largest cohort of patients evaluated to date, with cohorts stratified by SUD subtype and ICH versus non-ICH TBI, to highlight opportunities for post-injury care models to mitigate SUD risk. Methods: De-identified data from the TriNetX Research Network were used to select patients with TBI (n = 1,889,112) and define distinct cohorts based upon the presence (n = 420,868) or absence (n = 1,471,592) of ICH. Patients with previously diagnosed SUD before the date of TBI were excluded. Patient demographics and medical comorbidities were calculated for each group. The incidence of new SUD diagnosis over the lifetime and at 1-, 3-, and 5-years post-TBI were calculated and compared. Subtypes of SUD were defined and calculated based on the specific substance used. Propensity scores were calculated to create balanced matched ICH and non-ICH cohorts (n = 331,812 each) were used for comparisons of 5-year SUD incidence. Results: In the full TBI cohort, 5-year new SUD incidence was 4.2% overall, with nicotine (2.4%) and alcohol (1.1%) predominating, followed by cannabis (0.9%) and opioids (0.4%). Rates of SUDs increased over time, but attenuated beyond 5 years, with approximately 50% of those who would ultimately be diagnosed with SUD manifesting (lifetime) by 3 years post-TBI. After propensity matching, non-ICH TBI showed higher 5-year risk for any SUD (4.2% vs. 3.6%; risk difference −0.65%, p < 0.0001) and all subtypes (p < 0.05) except inhalants (p = 0.53). Conclusions: This largest-to-date analysis of new-onset SUD post-TBI demonstrates significantly higher rates of SUD in TBI patients; rates of nicotine, alcohol, cannabis, and opioid use disorders were most common. Non-ICH TBI patients demonstrated greater rates of SUD after injury than patients with ICH-associated TBI. Of patients suffering from TBI without ICH who would eventually be diagnosed with SUD, approximately 50% had obtained that diagnosis within 3 years of the injury. Taken together, these findings demonstrate the clinical need for routine SUD screening in post-TBI care, especially for 3 years post-injury. Such an intervention has the potential to significantly alleviate the public health burden and associated cost of care for TBI-associated substance use disorder patients.

Web-based large language models (LLMs) are increasingly used by patients for medical self-assessment, but their efficacy in spine imaging diagnostics remains underexplored. This study systematically evaluated 5 leading multimodal LLMs-Grok 2, Grok 3, Grok 4, ChatGPT, and Gemini-for detecting scoliosis and lumbar spinal stenosis across radiographs and MRI modalities. We assessed 171 full-length anterior-posterior radiographs (100 with scoliosis, 71 normal) and 200 axial T2-weighted lumbar spine MRIs (100 with severe stenosis, 100 normal) from public databases. Models were prompted without examples to identify pathology and quantify certainty (0%-100%). Analyses included McNemar's test for accuracy and ANOVA for confidence levels. In scoliosis detection, Grok 4 exhibited superior accuracy (0.942), followed by Gemini (0.912), Grok 2 (0.890), ChatGPT (0.643), and Grok 3 (0.637). For stenosis, Gemini performed best (0.600), then Grok 4 (0.575), ChatGPT (0.545), Grok 2 (0.500), and Grok 3 (0.450). All models sustained >70% mean certainty (SD < 5.3%) across pathologies. ChatGPT and Grok 3 demonstrated reduced confidence in erroneous scoliosis responses (P < 0.0001), while only ChatGPT did so for stenosis. Gemini reported elevated confidence in incorrect stenosis responses (P < 0.0001). LLMs perform highly in scoliosis detection but struggle to identify lumbar stenosis. ChatGPT's superior confidence calibration suggests enhanced reliability. Performance inconsistencies across model iterations (e.g., Grok 3 underperforming Grok 2) underscore the necessity for specialized medical imaging training. Although promising for patient education in simple spine conditions, substantial advancements in accuracy and confidence metrics are essential prior to clinical adoption or broad patient utilization.

Background: Alzheimer’s disease (AD) is a neurodegenerative disease that involves the accumulation of pathological tau protein. While the mechanisms behind tau accumulation are not well understood, they are pivotal to understanding the pathogenesis of AD. Because protein clearance in the brain is mediated in part by vasculature, we hypothesize that vasculature may direct the accumulation and distribution of tau pathology. In this study, we analyzed the spatial relationship between tau distribution and vasculature to elucidate vasculature’s role in tau clearance.Method: Post‐mortem human tissue samples (∼1 cm³) from the inferior temporal gyrus of 6 AD and 6 control donors were sliced at thicknesses of 500 µm to 1 mm. Tissue slices were prepared with tissue clearing. Cleared samples were stained with antibodies for tau (AT8), vasculature (GLUT1), nuclei (DAPI), and neurons (HUD). Images were collected using confocal fluorescence microscopy. A virtual reality tracing method was developed to segment individual blood vessels across large tissue volumes and 12‐21 individual vessels were isolated in each sample. Novel MATLAB scripts were developed to calculate the intensity of tau as a function of distance from the blood vessels’ surfaces.Result: In AD samples, tau intensity varied along the length of the blood vessel and with distance from the blood vessel surface. Regions with high tau accumulation on the blood vessel surface exhibited elevated tau accumulation in the surrounding tissue. In regions with low tau accumulation on the blood vessel surface, tau intensity was observed to decrease near the vessel. The influence of blood vessels on tau intensity dissipated as distance from the vessel surface increased, where tau intensity approached the bulk value (average tau intensity across the entire image). Compared to AD samples, control tissue showed little AT8+ tau labeling and the relationship between tau and distance from blood vessels was not apparent.Conclusion: These results indicate that pathological tau accumulation is spatially related to blood vessels in AD. In addition, these data imply that blood vessels contribute to tau clearance and that vascular tau clearance is disrupted in regions with high tau accumulation.