Emergency Patient Triage Improvement through a Retrieval-Augmented Generation Enhanced Large-Scale Language Model.

OBJECTIVES: Emergency medical triage is crucial for prioritizing patient care in emergency situations, yet its effectiveness can vary significantly based on the experience and training of the personnel involved. This study aims to evaluate the efficacy of integrating Retrieval Augmented Generation (RAG) with Large Language Models (LLMs), specifically OpenAI's GPT models, to standardize triage procedures and reduce variability in emergency care. METHODS: We created 100 simulated triage scenarios based on modified cases from the Japanese National Examination for Emergency Medical Technicians. These scenarios were processed by the RAG-enhanced LLMs, and the models were given patient vital signs, symptoms, and observations from emergency medical services (EMS) teams as inputs. The primary outcome was the accuracy of triage classifications, which was used to compare the performance of the RAG-enhanced LLMs with that of emergency medical technicians and emergency physicians. Secondary outcomes included the rates of under-triage and over-triage. RESULTS: The Generative Pre-trained Transformer 3.5 (GPT-3.5) with RAG model achieved a correct triage rate of 70%, significantly outperforming Emergency Medical Technicians (EMTs) with 35% and 38% correct rates, and emergency physicians with 50% and 47% correct rates (p < 0.05). Additionally, this model demonstrated a substantial reduction in under-triage rates to 8%, compared with 33% for GPT-3.5 without RAG, and 39% for GPT-4 without RAG. CONCLUSIONS: The integration of RAG with LLMs shows promise in improving the accuracy and consistency of medical assessments in emergency settings. Further validation in diverse medical settings with broader datasets is necessary to confirm the effectiveness and adaptability of these technologies in live environments.

Neurological recovery rate and predictive factors of incomplete AIS grade C spinal cord injury in the older aged population.

STUDY DESIGN: Retrospective cohort study. OBJECTIVES: To define the prognosis and predictive factors for neurological improvement in older patients with incomplete spinal cord injury (SCI) of American Spinal Injury Association Impairment Scale grade C (AIS-C). SETTINGS: Multi-institutions in Japan. METHODS: We included patients aged ≥65 years with traumatic SCI of AIS-C who were treated conservatively or surgically with >3 follow-up months. To identify factors related to neurological improvement, patients were divided into three groups according to their neurological status at the final follow-up, with univariate among-group comparisons of demographics, radiographic, and therapeutic factors. Significant variables were included in the multivariate logistic regression analysis. RESULTS: Overall, 296 older patients with SCI of AIS-C on admission were identified (average age: 75.2 years, average follow-up: 18.7 months). Among them, 190 (64.2%) patients improved to AIS-D and 21 (7.1%) patients improved to AIS-E at final follow-up. There were significant among-group differences in age (p = 0.026), body mass index (p = 0.007), status of pre-traumatic activities of daily living (ADL) (p = 0.037), and serum albumin concentrations (p = 0.011). Logistic regression analysis showed no significant differences in variables in the stratified group of patients who improved to AIS-D. Meanwhile, serum albumin was a significant variable in patients who improved to AIS-E (p = 0.026; OR: 6.20, pre-traumatic ADL was omitted due to data skewness). CONCLUSIONS: Most older patients with incomplete AIS-C SCI demonstrated at least 1 grade of neurological improvement. However, <10% of patients achieved complete recovery. Key predictors of complete recovery were high serum albumin levels on admission and independent pre-traumatic ADL. SPONSORSHIP: No funding was received for this study.

A nationwide multicenter study of the cost effectiveness of five leading drugs for pharmacological management of cervicobrachial symptoms.

BACKGROUND: Cervicobrachial pain frequently affects the quality of life (QOL) of the general public and has a significant economic impact on the health care systems of various countries. There are a number of treatment options for this disease, including widely-used drug therapy, but the effectiveness of each option is indeterminate, and there have been no published cost-effectiveness analysis studies so far. This prospective observational study aimed to examine the cost-effectiveness of drug treatment for cervicobrachial symptoms. METHODS: A 6-month medication regimen for each of five frequently-prescribed drugs for cervicobrachial symptoms was administered to 322 patients at 24 centers in Japan. Outcome measures, including of the EuroQol Group 5D, Short Form-8, and Visual Analog Scale (VAS), were investigated at baseline and every month thereafter. Incremental cost-effectiveness ratios (ICERs) of the drug cost to quality-adjusted life years (QALYs) were calculated. A stratified analysis of patient characteristics was also performed to identify baseline factors potentially affecting cost-effectiveness. RESULTS: The ICER of entire drug treatment for cervicobrachial symptoms was 7,491,640 yen. Compared with the reference willingness-to-pay, the ICER was assumed to not be cost-effective. A certain number of QALYs were gained during the first 3 months after the treatment intervention, but almost no QALYs were gained during the following 3 months. Stratified analysis showed that cost-effectiveness was extremely low for patients with high baseline VAS and high QOL. CONCLUSIONS: The available medications for cervicobrachial symptoms did not have excellent cost-effectiveness. Although a certain number of QALYs were gained during the first 3 months after medication, no QALYs were gained in the latter half of the study period, suggesting that it is not advisable to continue the medication needlessly. LEVEL OF EVIDENCE: II, prospective cohort study.

Posterior Decompression and Fusion for Cervical Spondylodiscitis due to Osteoradionecrosis after Head-and-Neck Cancer Radiotherapy: Two Case Reports.

Introduction: Cervical spondylodiscitis due to osteoradionecrosis (ORN) after head-and-neck cancer radiotherapy is a severe complication. However, there are few reports on the surgical treatment of this condition. Case Report: We report two cases of cervical spondylodiscitis due to ORN, which were successfully treated with posterior decompression and fusion. The first case was in a 73-year-old male patient with spondylodiscitis at C3-C5, due to ORN. A posterior fusion of the spine (C2-T1) was performed, and a biopsy was conducted at a site separate from the incision for fusion. The second case was in a 76-year-old female patient with spondylodiscitis due to C4-C7 ORN. Cervical posterior decompression and fusion (C2-Th2) were performed, and decompression (C5-6) was conducted through an incision separate from that for the fusion.An anterior approach was avoided in both cases because of radiation-induced tissue changes. For these two patients with cervical spondylodiscitis due to ORN, surgery resulted in an improvement of infection and neurological deficits by posterior spinal fusion, isolation from decompression or biopsy of the infected area, and antibiotic treatment. Conclusion: Posterior decompression and fusion are effective for spondylodiscitis in the cervical spine after head-and-neck radiotherapy, treating both infection and neurological deficits. Spinal fusion that avoids the level of the infected vertebral body and decompression from separate skin incision sites may prevent the spread of infection. An anterior approach should be avoided because the risk of esophageal perforation and posterior pharyngeal wall defects is high.

Machine Learning and Deep Learning in Spinal Injury: A Narrative Review of Algorithms in Diagnosis and Prognosis.

Spinal injuries, including cervical and thoracolumbar fractures, continue to be a major public health concern. Recent advancements in machine learning and deep learning technologies offer exciting prospects for improving both diagnostic and prognostic approaches in spinal injury care. This narrative review systematically explores the practical utility of these computational methods, with a focus on their application in imaging techniques such as computed tomography (CT) and magnetic resonance imaging (MRI), as well as in structured clinical data. Of the 39 studies included, 34 were focused on diagnostic applications, chiefly using deep learning to carry out tasks like vertebral fracture identification, differentiation between benign and malignant fractures, and AO fracture classification. The remaining five were prognostic, using machine learning to analyze parameters for predicting outcomes such as vertebral collapse and future fracture risk. This review highlights the potential benefit of machine learning and deep learning in spinal injury care, especially their roles in enhancing diagnostic capabilities, detailed fracture characterization, risk assessments, and individualized treatment planning.

Facet Joint Bridging in Adolescent-onset Adult Idiopathic Scoliosis with Thoracolumbar/Lumbar Curves.

OBJECTIVE: This study aims to comprehend the natural history of Adolescent idiopathic scoliosis (AIS) patients and determine risk factors for facet joint bridging in adolescent-onset adult idiopathic scoliosis with thoracolumbar/lumbar (TL/L) curves. METHODS: We included 50 patients with residual AIS with TL/L curves (3 males, 47 females; age 41.5±17.3 years, TL/L Cobb angle 59.4±11.8°). They were >20 years old and diagnosed with AIS during their adolescence. Radiographic parameters were measured, and facet joint bridging was defined from axial computed tomography (CT) images. RESULTS: The sagittal vertical axis (SVA) significantly increased with age (r=0.71, p<0.01). Coronal Cobb angle of the TL/L curve, L4 tilt, C7 translation, lumbar lordosis (LL), Pelvic incidence (PI) -LL, pelvic tilt (PT), and thoracolumbar kyphosis (TLK) were also correlated to age (p<0.05). There were significant differences in age, SVA, PI-LL, vertebral bridging, facet tropism, and apical vertebral rotation (AVR) between the facet joint bridging group (n=10) and the non-facet joint bridging group (n=40). In the multivariate logistic regression analysis, SVA, vertebral bridging, and AVR emerged as notable risk determinants for facet joint bridging. The threshold for facet joint bridging based on SVA was 2.1 cm (area under the curve [AUC]: 0.801; sensitivity = 90%, specificity = 65%). CONCLUSION: This research revealed that large SVA, the presence of vertebral bridging, and large AVR are associated with facet joint bridging in adolescent-onset adult idiopathic scoliosis patients with TL/L curves. The cut-off value for facet joint bridging based on SVA was 2.1cm.

A Novel Evaluation for Vertebral Artery Course Using 3D Magnetic Resonance Imaging with Computed Tomography -like Bone Contrast and Magnetic Resonance Angiography: A Proof of Concept Study.

OBJECTIVE: Vertebral artery (VA) injury poses a significant risk in cervical spine surgery, necessitating accurate preoperative assessment. This study aims to introduce and validate a novel approach that combines the Fast field echo that resembles a computed tomography using restricted echo spacing (FRACTURE) sequence with Time of Flight (TOF) Magnetic Resonance Angiography (MRA) for comprehensive evaluation of VA courses in the cervical spine. MATERIALS AND METHODS: A total of eight healthy volunteers and two patients participated in this study. The FRACTURE sequence provided high-resolution bone images of the cervical spine, while TOF MRA offered non-invasive vascular imaging. Fusion images were created by merging FRACTURE and MRA modalities to simultaneously visualize cervical spine structures and VA courses. Board-certified orthopedic spine surgeons independently evaluated images to assess the visibility of anatomical characteristics of the VA course by Likert-scale. RESULTS: The FRACTURE-MRA fusion images effectively depicted the extraosseous course of the VA at the craniovertebral junction, the intraosseous course of the VA at the craniovertebral junction, the VA entrance level to the transverse foramen, and the side-to-side asymmetry of bilateral VAs. Additionally, clinical cases demonstrated the utility of the proposed technique in identifying anomalies and guiding surgical interventions. CONCLUSIONS: The integration of the FRACTURE sequence and TOF MRA presents a promising methodology for the precise evaluation of VA courses in the cervical spine. This approach improves preoperative planning for cervical spine surgery with detailed anatomy and is a valuable alternative to conventional methods without contrast agents.

Conversion of T2-Weighted Magnetic Resonance Images of Cervical Spine Trauma to Short T1 Inversion Recovery (STIR) Images by Generative Adversarial Network.

INTRODUCTION: The short T1 inversion recovery (STIR) sequence is advantageous for visualizing ligamentous injuries, but the STIR sequence may be missing in some cases. The purpose of this study was to generate synthetic STIR images from MRI T2-weighted images (T2WI) of patients with cervical spine trauma using a generative adversarial network (GAN).  Methods: A total of 969 pairs of T2WI and STIR images were extracted from 79 patients with cervical spine trauma. The synthetic model was trained 100 times, and the performance of the model was evaluated with five-fold cross-validation.  Results: As for quantitative validation, the structural similarity score was 0.519±0.1 and the peak signal-to-noise ratio score was 19.37±1.9 dB. As for qualitative validation, the incorporation of synthetic STIR images generated by a GAN alongside T2WI substantially enhances sensitivity in the detection of interspinous ligament injuries, outperforming assessments reliant solely on T2WI. CONCLUSION: The GAN model can generate synthetic STIRs from T2 images of cervical spine trauma using image-to-image conversion techniques. The use of a combination of synthetic STIR images generated by a GAN and T2WI improves sensitivity in detecting interspinous ligament injuries compared to assessments that use only T2WI.

Multimodal Deep Learning-based Radiomics Approach for Predicting Surgical Outcomes in Patients with Cervical Ossification of the Posterior Longitudinal Ligament.

STUDY DESIGN: A retrospective analysis. OBJECTIVE: This research sought to develop a predictive model for surgical outcomes in patients with cervical ossification of the posterior longitudinal ligament (OPLL) using deep learning and machine learning (ML) techniques. SUMMARY OF BACKGROUND DATA: Determining surgical outcomes assists surgeons in communicating prognosis to patients and setting their expectations. Deep learning and ML are computational models that identify patterns from large datasets and make predictions. METHODS: Of the 482 patients, 288 patients were included in the analysis. A minimal clinically important difference (MCID) was defined as gain in Japanese Orthopaedic Association (JOA) score of 2.5 points or more. The predictive model for MCID achievement at 1 year post-surgery was constructed using patient background, clinical symptoms, and preoperative imaging features (x-ray, CT, MRI) analyzed via LightGBM and deep learning with RadImagenet. RESULTS: The median preoperative JOA score was 11.0 (IQR: 9.0-12.0), which significantly improved to 14.0 (IQR: 12.0-15.0) at 1 year after surgery (P < 0.001, Wilcoxon signed-rank test). The average improvement rate of the JOA score was 44.7%, and 60.1% of patients achieved the MCID. Our model exhibited an area under the receiver operating characteristic curve of 0.81 and the accuracy of 71.9% in predicting MCID at 1 year. Preoperative JOA score and certain preoperative imaging features were identified as the most significant factors in the predictive models. CONCLUSION: A predictive ML and deep learning model for surgical outcomes in OPLL patients is feasible, suggesting promising applications in spinal surgery. LEVEL OF EVIDENCE: 4.

Bone union-promoting effect of romosozumab in an ovariectomized rat posterolateral lumbar fusion model.

Spinal fixation surgery has been increasingly performed in patients with osteoporosis. Romosozumab, a drug that was introduced in Japan recently, is known to possibly promote bone healing. However, few studies have reported the therapeutic effects of romosozumab in clinical practice in Japan. Therefore, here, we investigated the effects of romosozumab dosage on bone fusion promotion using an ovariectomized rat spinal fusion model. Eight-week-old female Sprague-Dawley rats were matched by body weight and divided into three groups: 1.0 romosozumab (R) group (Evenity®, 25 mg/kg), 1/10R group (Evenity®, 2.5 mg/kg), and control (C) group (saline). Subcutaneous injections were administered twice a week for 8 weeks postoperatively. Computed tomography scans were performed every 2 weeks from the time of surgery till 8 weeks postoperatively. The mean fusion rates in terms of volume were significantly higher in the R groups [1/10R, 1.0R] than in the C group from 4 weeks postoperatively. The rate of increase was significantly higher in the 1.0R group from 4 weeks postoperatively and in the 1/10R group from 6 weeks postoperatively, than in the C group. The proportion of trabecular bone area was approximately 1.5 times higher in the R groups than in the C group. No significant differences were observed between the R groups. Our results suggest that romosozumab stimulates bone growth at the graft site, and similar effects were achieved at 1/10 of the standard dosage.

Safety and feasibility of intravenous administration of a single dose of allogenic-Muse cells to treat human cervical traumatic spinal cord injury: a clinical trial.

INTRODUCTION: Spinal cord injury (SCI) is a devastating injury and remains one of the largest medical and social burdens because of its intractable nature. According to the recent advances in stem cell biology, the possibility of spinal cord regeneration and functional restoration has been suggested by introducing appropriate stem cells. Multilineage-differentiating stress enduring (Muse) cells are a type of nontumorigenic endogenous reparative stem cell. The positive results of Muse cell transplantation for SCI was shown previously. As a first step for clinical application in human SCI, we conducted a clinical trial aiming to confirm the safety and feasibility of intravenously injected donor-Muse cells. METHODS: The study design of the current trial was a prospective, multicenter, nonrandomized, nonblinded, single-arm study. The clinical trial registration number was JRCT1080224764. Patients with a cervical SCI with a neurological level of injury C4 to C7 with the severity of modified Frankel classification B1 and B2 were included. A primary endpoint was set for safety and feasibility. Our protocol was approved by the PMDA, and the trial was funded by the Life Science Institute, Tokyo, Japan. The present clinical trial recruited 10 participants (8 males and 2 females) with an average age of 49.3 ± 21.2 years old. All 10 participants received a single dose of allogenic CL2020 (a total of 15 × 106 cells, 2.1-2.7 × 105 cells/kg of body weight), which is a Muse cell-based product produced from human mesenchymal stem cells, by an intravenous drip. RESULTS: There were two reported severe adverse events, both of which were determined to have no causal relationship with Muse cell treatment. The change in the ISNCSCI motor score, the activity of daily living and quality of life scores showed statistically significant improvements compared to those data at the time of CL2020 administration. CONCLUSION: In the present trial, no safety concerns were identified, and Muse cell product transplantation demonstrated good tolerability. Future clinical trials with appropriate study designs incorporating a control arm will clarify the definitive efficacy of single-dose allogenic Muse cell treatment with intravenous administration to treat SCI. TRIAL REGISTRATION: jRCT, JRCT1080224764. Registered 03 July 2019, https://jrct.niph.go.jp/latest-detail/jRCT1080224764 .

Impact of malnutrition on mortality and neurological recovery of older patients with spinal cord injury.

This retrospective cohort study established malnutrition's impact on mortality and neurological recovery of older patients with cervical spinal cord injury (SCI). It included patients aged ≥ 65 years with traumatic cervical SCI treated conservatively or surgically. The Geriatric Nutritional Risk Index was calculated to assess nutritional-related risk. Overall, 789 patients (mean follow-up: 20.1 months) were examined and 47 had major nutritional-related risks on admission. One-year mortality rate, median survival time, neurological recovery, and activities of daily living (ADL) at 1 year post-injury were compared between patients with major nutrition-related risk and matched controls selected using 1:2 propensity score matching to adjust for age, pre-traumatic neurological impairment, and activity. In the Kaplan-Meier analysis, the median survival times were 44.9 and 76.5 months for patients with major nutrition-related risk and matched controls, respectively (p = 0.015). Matched controls had more individuals with a neurological improvement of American Spinal Injury Association Impairment Scale ≥ 1 grade (p = 0.039) and independence in ADL at 1 year post-injury than patients with major nutrition-related risk (p < 0.05). In conclusion, 6% of older patients with cervical SCI had major nutrition-related risks; they showed a significantly higher 1 year mortality rate, shorter survival time, poorer neurological improvement, and lower ADL at 1 year post-injury than matched controls.

Early Versus Delayed Surgery for Elderly Traumatic Cervical Spinal Injury: A Nationwide Multicenter Study in Japan.

STUDY DESIGN: Retrospective multicenter study. OBJECTIVES: The effectiveness of early surgery for cervical spinal injury (CSI) has been demonstrated. However, whether early surgery improves outcomes in the elderly remains unclear. This study investigated whether early surgery for CSI in elderly affects complication rates and neurological outcomes. METHODS: This retrospective multicenter study included 462 patients. We included patients with traumatic acute cervical spinal cord injury aged ≥65 years who were treated surgically, whereas patients with American Spinal Injury Association (ASIA) Impairment Scale E, those with unknown operative procedures, and those waiting for surgery for >1 month were excluded. The minimum follow-up period was 6 months. Sixty-five patients (early group, 14.1%) underwent surgical treatment within 24 hours, whereas the remaining 397 patients (85.9%) underwent surgery on a standby basis (delayed group). The propensity score-matched cohorts of 63 cases were compared. RESULTS: Patients in the early group were significantly younger, had significantly more subaxial dislocations (and fractures), tetraplegia, significantly lower ASIA motor scores, and ambulatory abilities 6 months after injury. However, no significant differences in the rate of complications, ambulatory abilities, or ASIA Impairment Scale scores 6 months after injury were observed between the matched cohorts. At 6 months after injury, 61% of the patients in the early group (25% unsupported and 36% supported) and 53% of the patients in the delayed group (34% unsupported and 19% supported) were ambulatory. CONCLUSIONS: Early surgery is possible for CSI in elderly patients as the matched cohort reveals no significant difference in complication rates and neurological or ambulatory recovery between the early and delayed surgery groups.