European Spine Journal最新文献

Background: The degree of scoliosis is assessed through the Cobb angle, which quantifies severity and is measured by clinicians on radiographs. With the increasing adoption of artificial intelligence (AI) in clinical workflows, there is uncertainty as to whether large language models (LLMs) with image processing capabilities can streamline and improve spinal deformity classification. This study aims to assess the diagnostic capabilities of 4 leading LLMs: ChatGPT, Gemini, Perplexity and Grok in calculating Cobb angles from radiographs.

Methods: A cross-sectional analysis of 122 scoliosis patients was undertaken. Cobb angles were independently calculated using Horos software by a fellowship-trained radiologist, serving as the reference standard. All 122 radiographs were further uploaded to each of the 4 AI models to identify the type of scoliosis, generate a Cobb angle overlay and calculate the Cobb angle. Qualitative usability was assessed through pre-defined questions ranked on a Likert scale. Statistical tests included mean difference, paired t-tests and intraclass correlation coefficients.

Results: Gemini produced no calculated Cobb angles. ChatGPT failed to produce Cobb angles in 90 radiographs, and, even when Cobb angles were calculated, there were large errors (MAE 58.6° ± 45.9°). Both Perplexity and Grok generated estimates for all thoracolumbar cases, with mean differences of 18.8° (± 13.3°) and 24.2° (± 18.3°), respectively. None of the AI models successfully identified the S-shaped scoliosis cases. All AI models demonstrated a difference greater than the clinically accepted difference (≤ 10%).

Conclusion: This study concludes that current commercially available LLMs show limited accuracy in Cobb angle measurement. Whilst out of the 4 AI models assessed, Perplexity and Grok displayed the highest performance, no model displayed an acceptable clinical ability. These findings highlight the need for a dedicated and rigorous development of a spinal deformity AI tool before clinical integration of Cobb angle determination.

Purpose: Spinal tumors are rare and heterogeneous lesions of the central nervous system. Minimally invasive spinal surgery (MISS) offers several advantages over conventional open laminectomy, including reduced soft-tissue trauma, lower blood loss, and faster recovery. This study presents a 20-year, single-center experience with the minimally invasive paramedian transmuscular approach for spinal tumor resection, assessing its feasibility, safety, and effectiveness across all spinal levels.

Methods: A retrospective review was conducted of 61 patients who underwent minimally invasive resection of spinal tumors at a tertiary neurosurgical center. Demographic, radiological, operative, and postoperative data were analyzed. Intraoperative neuromonitoring was routinely used. Outcomes included extent of resection, complications, neurological status, and hospital stay.

Results: The mean age was 54.4 ± 13.5 years, with a predominance of females (70.5%). Most tumors were intradural extramedullary, primarily meningiomas and schwannomas. Gross-total resection was achieved in 85% of cases. The mean operative time was 148 min and mean blood loss 197 mL. Complications occurred in 9 patients, with clinically relevant cerebrospinal fluid-related events observed in 3.3%. New neurological deficits developed in 3.3% of patients. The mean hospital stay was 2.2 days. No perioperative mortality occurred, and tumor recurrence was observed in 4.9% during follow-up.

Conclusion: The minimally invasive paramedian transmuscular approach allows safe resection of selected spinal tumors, including junctional lesions, while achieving high resection rates and acceptable perioperative morbidity. These findings support its use as a feasible surgical option in appropriately selected patients.

Purpose: Discrepancies between clinical and imaging findings complicate accurate diagnoses of spinal levels in lumbar disorders. Extreme variations in the 12th rib length may be associated with cranial or caudal shifts of the lumbosacral plexus; however, the anatomical basis remains speculative because of limited in vivo visualization. This study assessed whether the conus medullaris position and L4 morphology-particularly the branch to the lumbosacral trunk (LST)-can be visualized using magnetic resonance imaging (MRI) and their correlations with the 12th rib length and symptom deviation.

Methods: This study included 127 patients who underwent lumbar disc herniation surgery and 25 cadaveric specimens. MRI and radiography were used to measure the conus medullaris level, LST branch visibility, and 12th rib length. Cadaveric dissections evaluated the LST branch thickness and its association with the rib length and plexus position.

Results: Patients with symptoms of caudal shift had shorter 12th ribs, a more cranially located conus medullaris, and more frequent visibility of the LST branch using MRI (all p < 0.001). LST branches visible with MRI were linked to shorter ribs and higher conus levels. The conus position positively correlated with the rib length (r = 0.36). A cadaveric analysis revealed a strong inverse correlation between the rib length and LST branch thickness (r = -0.86).

Conclusion: The 12th rib length, conus medullaris level, and L4 morphology reflect cranial-caudal shifts in neural structures and symptoms. These anatomical markers may be clinically useful predictors of level-specific symptoms and support more accurate surgical planning.