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.

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.

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.

Efficacy and safety of pararectus approach for the treatment of acetabular fractures: A systematic review and meta-analysis.

BACKGROUND: The pararectus (PR) approach combines the advantages of the second window of the ilioinguinal approach and the medial view of the modified Stoppa approach. However, it is unclear whether the PR approach is more effective or safer than the other approaches, as few clinical studies have compared the PR approach with the other approaches. The aim of this study was to provide a systematic review and meta-analysis comparing the PR approach with the other approaches for the treatment of acetabular fractures and to answer the following question: Are the results of the PR approach superior to those of the other approaches in terms of reduction quality, operative time, operative blood loss, complications, and clinical outcomes for treatment of acetabular fractures? PATIENTS AND METHODS: A systematic literature review was conducted using relevant original studies from various databases. Pooling of data was performed using RevMan software (version 5.3, Cochrane Collaboration, Oxford, UK). A p-value<0.05 was considered to be significant. We calculated the mean differences for continuous data and odds ratio for dichotomous data with 95% confidence intervals for each outcome. Statistical heterogeneity was assessed based on I2 using the standard χ2 test. RESULTS: Five studies were included in this meta-analysis. The findings demonstrated that operative blood loss was significantly lower in the PR approach than in the other approaches (p=0.04). There was no significant difference in the rate of anatomical reduction, the operative time, the rate of complications, and the rate of excellent or good clinical score between the PR approach and the other approaches. DISCUSSION: The PR approach provided lower operative blood loss, although there was no significant difference in reduction quality, operative time, complications, and excellent or good clinical score between the PR approach and the other approaches. LEVEL OF EVIDENCE: III.

Posterior Decompression and Fixation for Thoracic Spine Ossification: A 10-Year Follow-Up Study.

Ossification of the posterior longitudinal ligament of the thoracic spine (T-OPLL) causes symptoms including leg and back pain, and motor and sensory deficits. This study retrospectively reviewed 32 patients who initially underwent posterior decompression with instrumented fusion (PDF) for T-OPLL between 2001 and 2012, with 20 qualifying for the final analysis after applying exclusion criteria. Exclusions included unknown preoperative neurological findings, follow-up less than 10 years, or prior spinal surgeries at other levels. Outcomes were assessed using the Japanese Orthopedic Association (JOA) score, recovery rate, and kyphotic angle. The average preoperative JOA score of 3.6 improved to 7.4 at 1 year post-surgery and remained at 7.4 at 10 years, with a recovery rate of 52%. The kyphotic angle at T4-12 increased from 26 degrees preoperatively to 29 degrees postoperatively and to 37 degrees at 10 years. At the fused levels, the angle remained at 26 degrees immediately post-operation and increased to 32 degrees at 10 years. Forty percent of patients required additional surgery, primarily for conditions related to cervical OPLL, such as myelopathy, or lumbar OPLL, such as radiculopathy, or cauda equina syndrome. In conclusion, PDF effectively reduces T-OPLL symptoms over the long term, but the high rate of additional surgeries calls for careful patient follow-up.

Differences in Risk Factors for Decreased Cervical Lordosis after Multiple-Segment Laminoplasty for Cervical Spondylotic Myelopathy and Ossification of the Posterior Longitudinal Ligament: A Pilot Study.

STUDY DESIGN: Retrospective study. PURPOSE: To compare the radiographic risk factors for decreased cervical lordosis (CL) after laminoplasty, focusing on the difference between cervical spondylotic myelopathy (CSM) and cervical ossification of the posterior longitudinal ligament (C-OPLL). OVERVIEW OF LITERATURE: A few reports compared the risk factors for decreased CL between CSM and C-OPLL although these two pathologies have their characteristics. METHODS: This study included 50 patients with CSM and 39 with C-OPLL who underwent multi-segment laminoplasty. Decreased CL was defined as the difference between preoperative and 2-year postoperative neutral C2-7 Cobb angles. Radiographic parameters included preoperative neutral C2-7 Cobb angles, C2-7 sagittal vertical axis (SVA), T1 slope (T1S), dynamic extension reserve (DER), and range of motion. The radiographic risk factors were investigated for decreased CL in CSM and C-OPLL. Additionally, the Japanese Orthopedic Association (JOA) score was assessed preoperatively and 2 years postoperatively. RESULTS: C2-7 SVA (p =0.018) and DER (p =0.002) were significantly correlated with decreased CL in CSM, while C2-7 Cobb angle (p =0.012) and C2-7 SVA (p =0.028) were correlated with decreased CL in C-OPLL. Multiple linear regression analysis revealed that greater C2-7 SVA (B =0.22, p =0.026) and small DER (B =-0.53, p =0.002) were significantly associated with decreased CL in CSM. By contrast, greater C2-7 SVA (B =0.36, p =0.031) was significantly associated with decreased CL in C-OPLL. The JOA score significantly improved in both CSM and C-OPLL (p <0.001). CONCLUSIONS: C2-7 SVA was associated with a postoperative decreased CL in both CSM and C-OPLL, but DER was only associated with decreased CL in CSM. Risk factors for decreased CL slightly differed depending on the etiology of the condition.

Machine Learning Web Application for Predicting Functional Outcomes in Patients With Traumatic Spinal Cord Injury Following Inpatient Rehabilitation.

Accurately predicting functional outcomes in patients with spinal cord injury (SCI) helps clinicians set realistic functional recovery goals and improve the home environment after discharge. The present study aimed to develop and validate machine learning (ML) models to predict functional outcomes in patients with SCI and deploy the models within a web application. The study included data from the Japan Association of Rehabilitation Database from January 1, 1991, to December 31, 2015. Patients with SCI who were admitted to an SCI center or transferred to a participating post-acute rehabilitation hospital after receiving acute treatment were enrolled in this database. The primary outcome was functional ambulation at discharge from the rehabilitation hospital. The secondary outcome was the total motor Functional Independence Measure (FIM) score at discharge. We used binary classification models to predict whether functional ambulation was achieved, as well as regression models to predict total motor FIM scores at discharge. In the training dataset (70% random sample) using demographic characteristics and neurological and functional status as predictors, we built prediction performance matrices of multiple ML models and selected the best one for each outcome. We validated each model's predictive performance in the test dataset (the remaining 30%). Among the 4181 patients, 3827 were included in the prediction model for the total motor FIM score. The mean (standard deviation [SD]) age was 50.4 (18.7) years, and 3211 (83.9%) patients were male. There were 3122 patients included in the prediction model for functional ambulation. The CatBoost Classifier and regressor models showed the best performances in the training dataset. On the test dataset, the CatBoost Classifier had an area under the receiver operating characteristic curve of 0.8572 and an accuracy of 0.7769 for predicting functional ambulation. Likewise, the CatBoost Regressor performed well, with an R2 of 0.7859, a mean absolute error of 9.2957, and a root mean square error of 13.4846 for predicting the total motor FIM score. The final models were deployed in a web application to provide functional predictions. The application can be found at http://3.138.174.54:8501. In conclusion, our prediction models developed using ML successfully predicted functional outcomes in patients with SCI and were deployed in an open-access web application.

Mobile application for home exercise adherence in patients with knee osteoarthritis: A pilot study.

BACKGROUND: The adherence to home exercise is generally low despite its well-known effect on knee osteoarthritis. Therefore, we developed a home exercise application, LongLifeSupport, to provide patients with daily basic exercise videos and an automatic recording calendar. We hypothesized that this application would encourage patients to exercise and help maintain their motivation; this pilot study aimed to determine their exercise adherence rates. Using outcome measures, we also aimed to determine the effect of home exercise using this application and the factors for its continuation. METHODS: Twenty patients with knee osteoarthritis were included. The participants exercised for 12 weeks. Using pre- and post-tests, we examined their satisfaction with continuation (only in the post-test), Japanese knee osteoarthritis measure score, short physical performance battery score, bilateral knee extension muscle strength, and short test battery for locomotive syndrome. Furthermore, we investigated correlations between adherence rates and pretest scores of Japanese knee osteoarthritis measure and short test battery and between pretest scores and variations in Japanese knee osteoarthritis measure and short test battery. RESULTS: The mean adherence rate was 82.4%. The participants showed ease of continuation (100%) and significant improvements in the degree of knee pain, pain, and stiffness, and daily life conditions using the Japanese knee osteoarthritis measure score, total score, walk seconds, and chair stand seconds of the short physical performance battery, as well as the extension muscle strength of the right- and pain-side knee. No significant correlations were identified between the adherence rate and the pretest or variation. CONCLUSION: The adherence rate to the application was over 80%. Participants with knee osteoarthritis showed almost full satisfaction, reduced pain, and improved physical ability. Therefore, the use of this application provided a safe exercise program and maintained the exercise motivation of participants. Thus, it may be useful for unsupervised home exercise.

Automated fracture screening using an object detection algorithm on whole-body trauma computed tomography.

The emergency department is an environment with a potential risk for diagnostic errors during trauma care, particularly for fractures. Convolutional neural network (CNN) deep learning methods are now widely used in medicine because they improve diagnostic accuracy, decrease misinterpretation, and improve efficiency. In this study, we investigated whether automatic localization and classification using CNN could be applied to pelvic, rib, and spine fractures. We also examined whether this fracture detection algorithm could help physicians in fracture diagnosis. A total of 7664 whole-body CT axial slices (chest, abdomen, pelvis) from 200 patients were used. Sensitivity, precision, and F1-score were calculated to evaluate the performance of the CNN model. For the grouped mean values for pelvic, spine, or rib fractures, the sensitivity was 0.786, precision was 0.648, and F1-score was 0.711. Moreover, with CNN model assistance, surgeons showed improved sensitivity for detecting fractures and the time of reading and interpreting CT scans was reduced, especially for less experienced orthopedic surgeons. Application of the CNN model may lead to reductions in missed fractures from whole-body CT images and to faster workflows and improved patient care through efficient diagnosis in polytrauma patients.

Characteristics of stepladder fall injuries: a retrospective study.

BACKGROUND: Stepladders are used to work at heights in daily life, but their structure and usage differs from that of a ladder; stepladders can fold and stand without support, whereas ladders cannot fold and, thus, require support from other objects. We hypothesised that this difference made ladder and stepladder fall injuries to differ in characteristics. To clarify this hypothesis, we performed a retrospective cohort study on the stepladder fall injuries and compared their characteristics with that of ladder fall injuries. MATERIALS AND METHODS: We conducted a retrospective cohort study of injuries sustained from either ladder or stepladder falls. In this study, data were retrieved from the computerised database of Teikyo University Chiba Medical Center. Patients admitted to the orthopaedic department because of injury from ladder or stepladder fall were included. The following data were retrieved from the patient records: sex, age, height, body weight, body mass index (BMI), injury severity score (ISS), season, number of injury sites, details of injury and treatment option. RESULTS: One hundred thirty-two patients were included in this study. 101 patients were injured from stepladder falls, and 31 patients were injured from ladder falls. The number of females sustaining injuries due to a stepladder fall was significantly higher than those due to a ladder fall. The most frequent type of injury after fall from stepladder was fracture (48.9%), whereas the most frequent type of injury after fall from ladder was contusion/sprain (56.4%). The most frequently injured body part from stepladder fall was lower extremity (32.6%). In contrast, the most frequently injured body part due to a ladder fall was spine (27.3%). CONCLUSION: The current study found that the number of females sustaining injuries due to a stepladder fall was significantly higher than those due to a ladder fall. Furthermore, the most frequent body parts that needed surgery following a ladder fall injury were spine and upper extremity, whereas the most frequent body parts that needed surgery following a stepladder fall injury was lower extremity. Our study indicated that stepladder falls cause severe injuries and physical disability and can be a huge financial burden.

A case of hip joint septic arthritis due to haematogenous infection, which was misdiagnosed sternoclavicular joint septic arthritis as SAPHO syndrome.

Anterior chest pain associated with sternoclavicular arthritis has been considered a symptom specific to SAPHO syndrome. Differentiating aseptic arthritis of the SAPHO syndrome from infectious sternoclavicular joint (SCJ) arthritis is often difficult. We reported a 55-year-old woman with left hip joint and right SCJ pain. Diagnosis and treatment were delayed because she was misdiagnosed with SAPHO syndrome, and haematogenous infection spread to the hip joint septic arthritis. She was diagnosed promptly after presenting to our hospital and underwent early debridement, and her hip joint and SCJ was found to have almost no dysfunction. Differentiation between sternoclavicular arthritis of the SAPHO syndrome and sternoclavicular septic arthritis is clinically important and requires careful attention.

A new closed reduction technique using a traction table to treat a traumatic posterior dislocation of the hip joint.

Traumatic hip dislocations should be reduced gently and promptly to prevent complications such as osteonecrosis of the femoral head, heterotopic ossification, incomplete recovery of sciatic nerve injury, and fractures of the femoral neck. Many closed reduction manoeuvres have been reported; however, these manoeuvres require forceful and sustained traction, which places an enormous physical burden on surgeons. The aim of the present technical note is to describe a new reduction manoeuvre that uses a traction table to decrease the physical burden experienced by surgeons.