Use of artificial intelligence to identify data elements for The Japanese Orthopaedic Association National Registry from operative records.

BACKGROUND: The Japanese Orthopaedic Association National Registry (JOANR) was recently launched in Japan and is expected to improve the quality of medical care. However, surgeons must register ten detailed features for total hip arthroplasty, which is labor intensive. One possible solution is to use a system that automatically extracts information about the surgeries. Although it is not easy to extract features from an operative record consisting of free-text data, natural language processing has been used to extract features from operative records. This study aimed to evaluate the best natural language processing method for building a system that automatically detects some elements in the JOANR from the operative records of total hip arthroplasty. METHODS: We obtained operative records of total hip arthroplasty (n = 2574) in three hospitals and targeted two items: surgical approach and fixation technique. We compared the accuracy of three natural language processing methods: rule-based algorithms, machine learning, and bidirectional encoder representations from transformers (BERT). RESULTS: In the surgical approach task, the accuracy of BERT was superior to that of the rule-based algorithm (99.6% vs. 93.6%, p < 0.001), comparable to machine learning. In the fixation technique task, the accuracy of BERT was superior to the rule-based algorithm and machine learning (96% vs. 74%, p < 0.0001 and 94%, p = 0.0004). CONCLUSIONS: BERT is the most appropriate method for building a system that automatically detects the surgical approach and fixation technique.

Responsiveness of the Zurich Claudication Questionnaire, the Oswestry Disability Index, the Japanese Orthopaedic Association Back Pain Evaluation Questionnaire, the 8-Item Short Form Health Survey, and the Euroqol 5 dimensions 5 level in the assessment of patients with lumbar spinal stenosis.

PURPOSE: To assess the responsiveness of the Zurich Claudication Questionnaire (ZCQ), the Oswestry Disability Index, the Japanese Orthopaedic Association Back Pain Evaluation Questionnaire, the visual analog scale (VAS), the 8-Item Short Form Health Survey (SF-8), and the EuroQol 5 dimensions 5 level as methods of assessing outcomes of surgery for lumbar spinal stenosis. METHODS: We analyzed 218 patients who had undergone lumbar surgery for spinal stenosis and completed one year of follow-up. The internal responsiveness of each questionnaire and any domains was assessed by the effect size and standardized response mean. External responsiveness was assessed by the Spearman rank correlation coefficient and the receiver operating characteristics (ROC) curve. RESULTS: The most responsive assessments were "symptom severity" and "physical function" on the ZCQ, "walking ability" on the JOABPEQ, "leg pain" on the VAS, and "social function" on the JOABPEQ. The moderately responsive assessments were the physical component summary on the SF-8, the ODI, the EQ5D-5L, "low back pain" on the JOABPEQ, and "leg numbness" on the VAS. The least responsive assessments were "low back pain" on the VAS, "mental health" and "lumbar function" on the JOABPEQ, and the mental component summary on the SF-8. CONCLUSIONS: Because of its high responsiveness, "symptom severity" on the ZCQ is recommended as a primary tool for assessing outcome when designing prospective studies for lumbar spinal stenosis.

Cost-effectiveness of posterior lumbar interbody fusion in the Japanese universal health insurance system.

BACKGROUND: Globally, the cost-effectiveness of spinal surgery is becoming increasingly important. However, these data are limited to a few countries. The purpose of our study was to examine the cost/quality adjusted life year (cost/QALY) gained for posterior lumbar interbody fusion (PLIF) in the Japanese universal health insurance system. METHODS: Fifty five patients underwent PLIF for lumbar degenerative spinal canal stenosis between July 2013 and September 2015 was included. Effectiveness was measured using Euro QOL 5-dimension (EQ-5D), Short-Form 8 physical component summary (PCS), and visual analog scale (VAS). The cost was calculated from the perspective of the public healthcare payer. Effectiveness and cost were measured one year after surgery. QALYs were calculated by multiplying the utility value (EQ-5D) and life years. Only direct costs were included on the basis of actual reimbursements. Cost/QALY at a 5-year time horizon with a 2% discount rate was estimated. Sensitivity analysis was performed by varying the time horizon (2 years or 10 years). The exchange rate was defined as US $1 to Japanese 100 yen. RESULTS: Mean total cost one year after surgery was ï¿¥2,802,900 ($28029). Operative cost was ï¿¥1,779,700 ($17797). Mean gained score was 0.22 in EQ-5D, 10.3 in PCS, and -44 in VAS. Cost/QALY was ï¿¥2,697,500 ($26975). Sensitivity analysis demonstrated that cost/QALY at a 10-year time horizon was ï¿¥1,428,300 ($14283) and that cost/QALY at a 2-year time horizon was ï¿¥6,435,400 ($64354). CONCLUSIONS: Clinical outcomes after PLIF improved beyond minimum clinical improvement difference. Cost/QALY was below the widely-accepted benchmark (cost/QALY < $50000). PLIF could be regarded as cost-effective interventions.

Cost-effectiveness of lumbar fenestration surgery in the Japanese universal health insurance system.

BACKGROUND: Healthcare costs are a global concern, and cost-effectiveness analyses of interventions have become important. However, data regarding cost-effectiveness are limited to a few medical fields. The purpose of our study was to examine the Japanese universal health insurance system cost per quality-adjusted life year (QALY) for lumbar fenestration surgery. METHODS: Forty-eight patients who underwent fenestration for lumbar degenerative spinal canal stenosis between July 2013 and September 2015 were included. Effectiveness was evaluated by measuring the EuroQOL 5-dimension (EQ-5D), Short-Form 8 physical component summary (PCS), and visual analog scale (VAS). Cost was analyzed from the perspective of the public healthcare payer. Effectiveness and cost were measured 1 year after surgery. QALYs were calculated by multiplying the utility value (EQ-5D) and life years. Only direct costs based on actual reimbursements were included. Cost per QALY with a 5-year time horizon with a 2% discount rate was estimated. Sensitivity analysis was performed by varying the time horizon (2 years or 10 years). RESULTS: Mean total cost 1 year after fenestration surgery was 1,254,300 yen (standard deviation [SD], 430,000 yen; median, 1,172,300 yen). Operative cost was 406,800 yen (SD, 251,500 yen; median, 363,000 yen). Mean gained score was 0.21 for EQ-5D (SD, 0.18; median, 0.24), 11 for PCS (SD, 10; median, 12), and -43 for VAS (SD, 34; median, -38). Cost per QALY was 1,268,600 yen. Sensitivity analysis demonstrated that cost per QALY with a 10-year time horizon was 679,300 yen and that with a 2-year time horizon was 3,004,600 yen. CONCLUSIONS: Cost per QALY of lumbar fenestration with a 5-year time horizon was 1,268,600 yen (11,532 US dollar), which was below the widely accepted benchmark (cost per QALY <5,000,000-6,500,000 yen (50,000 US dollars)). Fenestration is a cost-effective intervention.

Circulating sclerostin and dickkopf-1 levels in ossification of the posterior longitudinal ligament of the spine.

Sclerostin and dickkopf-1(DKK1) are Wnt/ß-catenin signal antagonists that play an important role in bone formation. Ossification of the posterior longitudinal ligament (OPLL) of the spine is characterized by pathological ectopic ossification of the posterior longitudinal ligament and ankylosing spinal hyperostosis. The aims of this study were to evaluate serum sclerostin and DKK1 levels in persons with OPLL and to identify its relationship with bone metabolism and bone mass in persons with OPLL. This was a case-control study, and 78 patients with OPLL were compared with 39 age- and sex-matched volunteers without OPLL. We analyzed the relationship with calciotropic hormones, bone turnover markers, OPLL localization, number of ossified vertebrae, and bone mineral density of total hip (TH-BMD). Serum sclerostin levels in men with OPLL were significantly higher than in men in the control group (control group: mean = 45.3 pmol/L; OPLL group: mean = 75.7 pmol/L; P = 0.002). Age and sclerostin levels were positively correlated in men with OPLL (r = 0.43; P = 0.002). Serum sclerostin levels in men with OPLL had a positive correlation with TH-BMD Z-score (r = 0.511; P = 0.011, n = 30). There was a strong negative correlation between serum sclerostin levels and serum DKK1 levels in men with OPLL (r = -0.506; P < 0.001). Bone and mineral metabolism in OPLL differs between men and women. In men with OPLL, systemic secretion of sclerostin increases with advancing age and with higher bone mass. These two Wnt/ß-catenin signal antagonists have the opposite effect in persons with OPLL, and higher serum sclerostin levels are counterbalanced by underproduction of DKK1.

Factors affecting postoperative activities of daily living in patients with osteoporotic vertebral collapse with neurological deficits.

Surgical treatment of osteoporotic vertebral collapse (OVC) with neurological deficits presents significant clinical challenges because some patients have fragile bones and often have medical comorbidities, which affect the severity of osteoporosis. We hypothesized that clinical results of surgery in these patients depend on the extent of medical comorbidities that induce secondary osteoporosis. The aim of this study is to examine the effects of medical history and comorbidities on surgical outcomes for these patients, along with the factors that predict postoperative function in activities of daily living (ADL). We retrospectively reviewed data for 88 patients with OVC and neurological deficits who underwent surgery. We assessed clinical results regarding neurological deficits and function in ADL. The presence or absence of comorbidities responsible for secondary osteoporosis and treatments or medical events that affect bone metabolism were examined. We performed statistical analysis to examine prognostic factors for postoperative function in ADL. Of 88 patients, the distributions of comorbidities, treatment, and events in medical history were as follows: hypertension, 57 patients (64.8%); chronic kidney disease (CKD) stage 3 or 4, 32 (36.4%); diabetes mellitus, 16 (18.2%); liver dysfunction, 11 (12.5%); cardiovascular disease, 10 (11.4%); rheumatoid arthritis, 9 (10.2%); and glucocorticoid intake, 8 (9.1%). Twenty-five patients (28.4%) represented poor postoperative ADL (chair-bound or bed-bound), and 11 of 25 patients with poor postoperative ADL represented full neurological recovery. Multivariate analysis revealed decreased estimated glomerular filtration rate (odds ratio 0.96; 95% confidence interval 0.93-0.99; p = 0.005) and a high serum alkaline phosphatase (ALP) level (odds ratio 1.01; 95% CI 1.00-1.02; p = 0.01) were strong predictive factors for poor postoperative function in ADL. The majority of patients with poor postoperative function in ADL had advanced CKD with a disorder of bone metabolism as well as bone fragility.

Analysis of mechanical failure associated with reoperation in spinal fusion to the sacrum in adult spinal deformity.

BACKGROUND: Long fusion to the sacrum has been demonstrated to increase the mechanical failure rate after adult spinal deformity (ASD) surgery, and these mechanical failures are the most common etiology for reoperation. The purpose of this study was to determine the incidence and risk factors for mechanical failure associated with reoperation after spinal fusion to the sacrum in ASD. METHODS: The study included 76 patients with ASD who underwent spinal fusion surgery including the sacrum at a single institution between 2005 and 2010. The inclusion criteria were a minimum age of 20 years and fusion of ≥ 5 levels. The terminal event was defined as either the first reoperation for mechanical failure or a minimum of 2 years following surgery in patients who did not undergo reoperation. RESULTS: The cumulative reoperation rate for mechanical failure was 37 % (n = 28). The procedure survival rate was 79 % at 1 year and 72 % at 2 years. Mechanical failures consisted of proximal junctional complications in 16 patients and pseudarthrosis in 12 patients. Proximal junctional kyphosis (PJK) was the most frequent cause (n = 15), and seven patients were diagnosed with fractures at the UIV or one level above the UIV. Multivariate analysis identified the following as independent factors predicting mechanical failure: three or more comorbidities, smoking, and a preoperative sagittal vertical axis of >95 mm. SRS-22r and ODI scores were lower in patients with mechanical failure. CONCLUSION: Overall, 37 % of the patients who underwent ASD surgery involving the sacrum required reoperation for mechanical failure. The most frequent form of mechanical failure associated with reoperation was surgical PJK. Significant risk factors for mechanical failure included medical comorbidities, smoking, and severe preoperative sagittal imbalance. Critical mechanical failure may have a negative influence on health status.

3D morphometric analysis of laminae and facet joints in patients with degenerative spondylolisthesis.

OBJECTIVE: To clarify the three-dimensional (3D) morphometric characteristics of the spine in patients with degenerative spondylolisthesis (DS). METHODS: 3D morphometric analyses of laminae and facets were performed and compared for a DS group, an age-matched spinal canal stenosis (LCS) group, and a control group of young persons without spinal disease. 3D facet sagittal angles (3D-FSAs), 3D facet axial angle (3D-FAAs), and 3D-FAA tropism at L3 and at L4 were measured by extracting the 3D inferior articular process. The 3D lamina inclination angles (3D-LIAs) of L3 and L4 were also measured by extracting the ventral surface of the laminae. RESULTS: The 3D-FSAs at L4 in the DS group were significantly higher than for the other groups, but the difference in 3D-FSAs at L3 was not statistically significant among the groups. The 3D-FAAs at L4 in the DS group were significantly lower than in the control group. There was no significant difference in other factors. CONCLUSIONS: 3D morphometric analysis clarified that DS is significantly correlated with horizontalization (higher 3D-FSA), but is not correlated with sagittalization (lower 3D-FAA) and tropism (3D-FAA tropism) of facet joints or horizontalization of laminae (3D-LIA). There were no morphometric characteristics at the cranial adjacent segment of DS.

Long fusion from sacrum to thoracic spine for adult spinal deformity with sagittal imbalance: upper versus lower thoracic spine as site of upper instrumented vertebra.

OBJECT: Despite increasing numbers of patients with adult spinal deformity, it is unclear how to select the optimal upper instrumented vertebra (UIV) in long fusion surgery for these patients. The purpose of this study was to compare the use of vertebrae in the upper thoracic (UT) versus lower thoracic (LT) spine as the upper instrumented vertebra in long fusion surgery for adult spinal deformity. METHODS: Patients who underwent fusion from the sacrum to the thoracic spine for adult spinal deformity with sagittal imbalance at a single medical center were studied. The patients with a sagittal vertical axis (SVA) ≥ 40 mm who had radiographs and completed the 12-item Short-Form Health Survey (SF-12) preoperatively and at final follow-up (≥ 2 years postoperatively) were included. RESULTS: Eighty patients (mean age of 61.1 ± 10.9 years; 69 women and 11 men) met the inclusion criteria. There were 31 patients in the UT group and 49 patients in the LT group. The mean follow-up period was 3.6 ± 1.6 years. The physical component summary (PCS) score of the SF-12 significantly improved from the preoperative assessment to final follow-up in each group (UT, 34 to 41; LT, 29 to 37; p = 0.001). This improvement reached the minimum clinically important difference in both groups. There was no significant difference in PCS score improvement between the 2 groups (p = 0.8). The UT group had significantly greater preoperative lumbar lordosis (28° vs 18°, p = 0.03) and greater thoracic kyphosis (36° vs 18°, p = 0.001). After surgery, there was no significant difference in lumbar lordosis or thoracic kyphosis. The UT group had significantly greater postoperative cervicothoracic kyphosis (20° vs 11°, p = 0.009). The UT group tended to maintain a smaller positive SVA (51 vs 73 mm, p = 0.08) and smaller T-1 spinopelvic inclination (-2.6° vs 0.6°, p = 0.06). The LT group tended to have more proximal junctional kyphosis (PJK), although the difference did not reach statistical significance. Radiographic PJK was 32% in the UT group and 41% in the LT group (p = 0.4). Surgical PJK was 6.4% in the UT group and 10% in the LT group (p = 0.6). CONCLUSIONS: Both the UT and LT groups demonstrated significant improvement in clinical and radiographic outcomes. A significant difference was not observed in improvement of clinical outcomes between the 2 groups.

New classification system for ossification of the posterior longitudinal ligament using CT images.

BACKGROUND: Ossification of the posterior longitudinal ligament (OPLL) is most frequently seen in the cervical spine. The types of cervical OPLL are classified into continuous, mixed, segmental, and other based on plain lateral X-ray. Computed tomography (CT) imaging is often used in clinical practice for evaluating ossified lesions as it can detect their precise location, size, and shape. However, to date, no CT classification of OPLL lesions has been proposed. METHODS: One hundred and forty-four patients diagnosed with cervical OPLL by plain radiograph were included in this study. Sagittal and axial CT images of the cervical spine were obtained. We propose three classification systems: A, B, and axial. Classification A comprises two lesion types: bridge and nonbridge. Classification B requires examiners to describe all vertebral and intervertebral levels where OPLL exits in the cervical spine. Axial classification comprises central and lateral lesions identified on axial CT images. Seven observers evaluated CT images using this classification system, and intra- and interrater reliability were examined. RESULTS: Averaged Fleiss' kappa coefficient of interrater agreement was 0.43 ± 0.26 among the seven observers, averaged intrarater reliability for the existence of OPLL was 72.4 ± 8.8% [95% confidence interval (CI) 67.5-76.8]. Fifty-four patients (37.5%) had the bridge type and 90 the nonbridge type according to Classification A; 102 (70.8%) had central and 42 (29.2%) lateral OPLL in the axial classification. Four representative cases defined according to the three classification types are reported here. CONCLUSION: Subcommittee members of the Investigation Committee on the Ossification of the Spinal Ligaments of the Japanese Ministry of Public Health and Welfare propose three new classification systems of cervical OPLL based on CT imaging: A, B, and axial.

Postoperative Changes of Spinopelvic Sagittal Parameters After Cervical Laminoplasty for Cervical Spondylotic Myelopathy.

STUDY DESIGN: A retrospective analysis of prospectively collected data. OBJECTIVE: To investigate postoperative changes of spinopelvic sagittal parameters after laminoplasty for cervical spondylotic myelopathy (CSM) accompanying postoperative cervical kyphotic deformity or cervical regional sagittal imbalance. SUMMARY OF BACKGROUND DATA: To the best of our knowledge, no study has been reported concerning postoperative changes of spinopelvic sagittal parameters accompanying postoperative deterioration of cervical sagittal alignment or balance after cervical laminoplasty. METHODS: Forty-five CSM patients without preoperative cervical kyphosis who underwent laminoplasty were included. None of the 45 patients had a medical history of previous spine surgery, hip joint surgery, or knee joint surgery. The patients were divided into 2 groups (kyphosis and lordosis groups) according to postoperative C2-7 angle, and they were also divided into 2 other groups (imbalance and balance groups) according to postoperative C1-7 sagittal vertical axis. Postoperative changes (Δ) of T1 slope (T1S), thoracic kyphosis, thoracolumbar kyphosis (TLK), lumbar lordosis (LL), Pelvic tilt, and C7 sagittal vertical axis were measured comparing lateral radiographs of the whole spine in the standing position taken at 1 year postoperatively with those before surgery. RESULTS: Both T1S and TLK significantly decreased after cervical laminoplasty in the kyphosis group compared with the lordosis group. On the other hand, both T1S and TLK increased significantly, and LL significantly decreased after surgery in the imbalance group compared with the balance group. CONCLUSIONS: At 1 year after laminoplasty for CSM, both T1S and TLK significantly decreased accompanying postoperative cervical kyphotic deformity as a compensatory action for postoperative cervical kyphosis to maintain the global sagittal balance of the spine, whereas both T1S and TLK increased significantly, and LL significantly decreased accompanying postoperative cervical reginal sagittal imbalance which resulted in postoperative forward inclination of the whole spine.

Risk Factor Analysis of Surgery-related Complications in Primary Thoracic Spine Surgery for Degenerative Diseases and Characteristics of the Patients Also Undergoing Surgery on the Cervical and/or Lumbar Spine.

STUDY DESIGN: A retrospective cohort study using prospectively collected data. OBJECTIVE: This study primarily aimed to investigate the risk factors for surgery-related complications in primary thoracic spine surgery for degenerative diseases using a surgeon-maintained database. The secondary purpose was to elucidate the characteristics of surgically treated thoracic myelopathy that also required cervical and/or lumbar spine surgery in the study period. SUMMARY OF BACKGROUND DATA: Few studies reported surgical complications and the feature of tandem spinal stenosis in thoracic myelopathy in detail because of their rarity. MATERIALS AND METHODS: This study included 840 thoracic myelopathy patients undergoing primary surgery for degenerative diseases from 2012 to 2021, investigating the effects of diseases, surgical procedures, and patient demographics on postoperative neurological deterioration, dural tear, dural leakage, surgical-site infection, and postoperative hematoma. In thoracic myelopathy patients who were surgically treated and also undergoing cervical and/or lumbar surgery, we investigated the proportion, the effects of diseases, and the order and intervals between surgeries. RESULTS: Multivariate logistic regression revealed that significant risk factors ( P <0.05) for postoperative neurological deterioration were intervertebral disk herniation [odds ratio (OR): 4.59, 95% confidence interval (CI): 1.32-16.0) and degenerative spondylolisthesis (OR: 11.1, 95% CI: 2.15-57.5). Ossification of the ligamentum flavum (OR: 4.12, 95% CI: 1.92-8.86), anterior spinal fusion (OR: 41.2, 95% CI: 4.70-361), and circumferential decompression via a posterior approach (OR: 30.5, 95% CI: 2.27-410) were risk factors for dural tear. In thoracic myelopathy patients surgically treated, 37.0% also underwent degenerative cervical and/or lumbar surgery. CONCLUSIONS: Pathologies involving anterior decompression and instability increased the risk of postoperative neurological deterioration. The risk of dural tear was increased when dura mater adhesions were likely to be directly operated upon. It should be recognized that a relatively high proportion (37.0%) of surgically treated thoracic myelopathy patients also underwent cervical and/or lumbar surgery.

Appropriate pelvic obliquity measurement method in Lenke type 5C adolescent idiopathic scoliosis: a preliminary study.

Background: Although pelvic obliquity (PO) is a risk factor for postoperative coronal decompensation in corrective surgery in adolescent idiopathic scoliosis (AIS), especially Lenke 5C, methods of measuring PO are controversial. This study aimed to establish an appropriate measurement method using multiplanar reconstructed computed tomography (MPR-CT) images instead of standing posteroanterior (PA) whole-spine radiographs to evaluate PO in patients with Lenke 5C AIS. Methods: This study was a retrospective cross-sectional study. Twenty-five patients who underwent corrective surgery for AIS in Osaka University Hospital from August 2014 to February 2023 were included. Cobb angle, L5 tilt, C7 plumb line to center sacral vertebral line (C7PL-CSVL), and leg length discrepancy (LLD) were measured on standing PA whole-spine radiographs preoperatively. Sacral obliquity (SO), the slope of the upper endplate of S1, and iliac obliquity (IO), the tilt of the line connecting the iliac crests, were measured on standing PA whole-spine radiographs and MPR-CT (SO/IO-X-ray, SO/IO-CT, respectively). S1 angle and S2 angle were measured on CT. Results: The mean age of the patients was 18.7±3.9 years and all of them were females. SO-X-ray and SO-CT were larger than IO-X-ray and IO-CT, respectively. SO-X-ray was highly correlated with SO-CT (r=0.838, P<0.001). L5 tilt had higher correlation with SO-CT (r=0.884, P<0.001) than with SO-X-ray (r=0.726, P=0.001) and IO-CT (r=0.550, P=0.22). L5 tilt was correlated poorly with IO-X-ray (r=0.104, P=0.69). The S1 angle was 4.5±3.5° meanwhile the S2 angle was 1.2±2.1°, the sacral deformity was mainly due to the S1 vertebral wedging. Conclusions: Given the asymmetric sacral morphology, SO is more appropriate pelvic parameter than IO to represent the sacral tilt of Lenke 5C AIS, especially when measured using CT images to overcome the poor visibility on PA whole-spine radiographs.

A Comparison Between GPT-3.5, GPT-4, and GPT-4V: Can the Large Language Model (ChatGPT) Pass the Japanese Board of Orthopaedic Surgery Examination?

Introduction Recently, large-scale language models, such as ChatGPT (OpenAI, San Francisco, CA), have evolved. These models are designed to think and act like humans and possess a broad range of specialized knowledge. GPT-3.5 was reported to be at a level of passing the United States Medical Licensing Examination. Its capabilities continue to evolve, and in October 2023, GPT-4V became available as a model capable of image recognition. Therefore, it is important to know the current performance of these models because they will be soon incorporated into medical practice. We aimed to evaluate the performance of ChatGPT in the field of orthopedic surgery. Methods We used three years' worth of Japanese Board of Orthopaedic Surgery Examinations (JBOSE) conducted in 2021, 2022, and 2023. Questions and their multiple-choice answers were used in their original Japanese form, as was the official examination rubric. We inputted these questions into three versions of ChatGPT: GPT-3.5, GPT-4, and GPT-4V. For image-based questions, we inputted only textual statements for GPT-3.5 and GPT-4, and both image and textual statements for GPT-4V. As the minimum scoring rate acquired to pass is not officially disclosed, it was calculated using publicly available data. Results The estimated minimum scoring rate acquired to pass was calculated as 50.1% (43.7-53.8%). For GPT-4, even when answering all questions, including the image-based ones, the percentage of correct answers was 59% (55-61%) and GPT-4 was able to achieve the passing line. When excluding image-based questions, the score reached 67% (63-73%). For GPT-3.5, the percentage was limited to 30% (28-32%), and this version could not pass the examination. There was a significant difference in the performance between GPT-4 and GPT-3.5 (p < 0.001). For image-based questions, the percentage of correct answers was 25% in GPT-3.5, 38% in GPT-4, and 38% in GPT-4V. There was no significant difference in the performance for image-based questions between GPT-4 and GPT-4V. Conclusions ChatGPT had enough performance to pass the orthopedic specialist examination. After adding further training data such as images, ChatGPT is expected to be applied to the orthopedics field.

Development of patient-reported outcome for adult spinal deformity: validation study.

Adult spinal deformity (ASD) is a complex condition that combines scoliosis, kyphosis, pain, and postoperative range of motion limitation. The lack of a scale that can successfully capture this complex condition is a clinical challenge. We aimed to develop a disease-specific scale for ASD. The study included 106 patients (mean age; 68 years, 89 women) with ASD. We selected 29 questions that could be useful in assessing ASD and asked the patients to answer them. The factor analysis found two factors: the main symptom and the collateral symptom. The main symptom consisted of 10 questions and assessed activity of daily living (ADL), pain, and appearance. The collateral symptom consisted of five questions to assess ADL due to range of motion limitation. Cronbach's alpha was 0.90 and 0.84, respectively. The Spearman's correlation coefficient between the change of main symptom and satisfaction was 0.48 (p < 0.001). The effect size of Cohen's d for comparison between preoperative and postoperative scores was 1.09 in the main symptom and 0.65 in the collateral symptom. In conclusion, we have developed a validated disease-specific scale for ASD that can simultaneously evaluate the benefits and limitations of ASD surgery with enough responsiveness in clinical practice.

Identification of unapproved orthopedic regenerative medicine: Usefulness of the Act on Safety of Regenerative Medicine.

In Japan, the Act on Safety of Regenerative Medicine regulates unapproved regenerative medicine. Other nations market regenerative medicine products, bypassing regulatory approval. To identify unapproved orthopedic regenerative medicine, we have used data based on the Act. Platelet-rich plasma was often used. The common target was the knee. Prices averaged $2,490.

Automated entry of paper-based patient-reported outcomes: Applying deep learning to the Japanese orthopaedic association back pain evaluation questionnaire.

BACKGROUND: Health-related patient-reported outcomes (HR-PROs) are crucial for assessing the quality of life among individuals experiencing low back pain. However, manual data entry from paper forms, while convenient for patients, imposes a considerable tallying burden on collectors. In this study, we developed a deep learning (DL) model capable of automatically reading these paper forms. METHODS: We employed the Japanese Orthopaedic Association Back Pain Evaluation Questionnaire, a globally recognized assessment tool for low back pain. The questionnaire comprised 25 low back pain-related multiple-choice questions and three pain-related visual analog scales (VASs). We collected 1305 forms from an academic medical center as the training set, and 483 forms from a community medical center as the test set. The performance of our DL model for multiple-choice questions was evaluated using accuracy as a categorical classification task. The performance for VASs was evaluated using the correlation coefficient and absolute error as regression tasks. RESULT: In external validation, the mean accuracy of the categorical questions was 0.997. When outputs for categorical questions with low probability (threshold: 0.9996) were excluded, the accuracy reached 1.000 for the remaining 65 % of questions. Regarding the VASs, the average of the correlation coefficients was 0.989, with the mean absolute error being 0.25. CONCLUSION: Our DL model demonstrated remarkable accuracy and correlation coefficients when automatic reading paper-based HR-PROs during external validation.

Surgical treatment for osteoporotic vertebral collapse with neurological deficits: retrospective comparative study of three procedures--anterior surgery versus posterior spinal shorting osteotomy versus posterior spinal fusion using vertebroplasty.

PURPOSE: In general, osteoporotic vertebral collapse (OVC) with neurological deficits requires sufficient decompression of neural tissues to restore function level in activities of daily living (ADL). However, it remains unclear as to which procedure provides better neurological recovery. The primary purpose of this study was to compare neurological recovery among three typical procedures for OVC with neurological deficits. Secondary purpose was to compare postoperative ADL function. METHODS: We retrospectively reviewed data for 88 patients (29 men and 59 women) with OVC and neurological deficits who underwent surgery. Three typical kinds of surgical procedures with different decompression methods were used: (1) anterior direct neural decompression and reconstruction (AR group: 27 patients), (2) posterior spinal shorting osteotomy with direct neural decompression (PS group: 36 patients), and (3) posterior indirect neural decompression and short-segment spinal fusion combined with vertebroplasty (VP group: 25 patients). We examined clinical results regarding neurological deficits and function level in ADL and radiological results. RESULTS: The mean improvement rates for neurological deficits and ADL function level were 60.1 and 55.0%, respectively. There were no significant differences among three groups in improvement rates for neurological deficits or ADL function level. The VP group had a significantly lower estimated mean blood loss (338 mL) and mean duration of surgery (229 min) than both the AR and PS groups (p < 0.001). CONCLUSION: Direct neural decompression is not always necessary, and the majority of patients can be treated with a less-invasive procedure such as short-segment posterior spinal fusion with indirect decompression combined with vertebroplasty. The high-priority issue is careful evaluation of patients' general health and osteoporosis severity, so that the surgeon can choose the procedure best suited for each patient.

Is there a difference in range of motion, neck pain, and outcomes in patients with ossification of posterior longitudinal ligament versus those with cervical spondylosis, treated with plated laminoplasty?

OBJECT: There are little data on the effects of plated, or plate-only, open-door laminoplasty on cervical range of motion (ROM), neck pain, and clinical outcomes. The purpose of this study was to compare ROM after a plated laminoplasty in patients with ossification of posterior longitudinal ligament (OPLL) versus those with cervical spondylotic myelopathy (CSM) and to correlate ROM with postoperative neck pain and neurological outcomes. METHODS: The authors retrospectively compared patients with a diagnosis of cervical stenosis due to either OPLL or CSM who had been treated with plated laminoplasty in the period from 2007 to 2012 at the University of California, San Francisco. Clinical outcomes were measured using the modified Japanese Orthopaedic Association (mJOA) scale and neck visual analog scale (VAS). Radiographic outcomes included assessment of changes in the C2-7 Cobb angle at flexion and extension, ROM at C2-7, and ROM of proximal and distal segments adjacent to the plated lamina. RESULTS: Sixty patients (40 men and 20 women) with an average age of 63.1 ± 10.9 years were included in the study. Forty-one patients had degenerative CSM and 19 patients had OPLL. The mean follow-up period was 20.9 ± 13.1 months. The mean mJOA score significantly improved in both the CSM and the OPLL groups (12.8 to 14.5, p < 0.01; and 13.2 to 14.2, respectively; p = 0.04). In the CSM group, the mean VAS neck score significantly improved from 4.2 to 2.6 after surgery (p = 0.01), but this improvement did not reach the minimum clinically important difference (MCID). Neither was there significant improvement in the VAS neck score in the OPLL group (3.6 to 3.1, p = 0.17). In the CSM group, ROM at C2-7 significantly decreased from 32.7° before surgery to 24.4° after surgery (p < 0.01). In the OPLL group, ROM at C2-7 significantly decreased from 34.4° to 20.8° (p < 0.01). In the CSM group, the change in the VAS neck score significantly correlated with the change in the flexion angle (r = - 0.31) and the extension angle (r = - 0.37); however, it did not correlate with the change in ROM at C2-7 (r = - 0.1). In the OPLL group, the change in the VAS neck score did not correlate with the change in the flexion angle (r = 0.03), the extension angle (r = - 0.17), or the ROM at C2-7 (r = - 0.28). The OPLL group had a significantly greater loss of ROM after surgery than did the CSM group (p = 0.04). There was no significant correlation between the change in ROM and the mJOA score in either group. CONCLUSIONS: Plated laminoplasty in patients with either OPLL or CSM decreases cervical ROM, especially in the extension angle. Among patients who have undergone laminoplasty, those with OPLL lose more ROM than do those with CSM. No correlation was observed between neck pain and ROM in either group. Neither group had a change in neck pain that reached the MCID following laminoplasty. Both groups improved in neurological function and outcomes.

Development of a system to assess the two- and three-dimensional bone mineral density of the lumbar vertebrae from clinical quantitative CT images.

This study developed a system to quantify the lumbar spine's bone mineral density (BMD) in two and three dimensions for osteoporosis screening using quantitative CT images. Measuring the two-dimensional BMD could reproduce the BMD measurement performed in dual-energy X-ray absorptiometry, and an accurate diagnosis of osteoporosis was possible. PURPOSE: To date, the assessment of bone mineral density (BMD) using CT images has been made in three dimensions, leading to errors in detecting osteoporosis based on the two-dimensional assessments of BMD using dual-energy X-ray absorptiometry (DXA-BMD). Herein, we aimed to develop a system that measures two- and three-dimensional lumbar BMD from quantitative CT images and validated the accuracy of the system in diagnosing osteoporosis with regard to the DXA classification. METHODS: Fifty-nine pairs of spinal CT and DXA images were analyzed. First, the three-dimensional BMD was measured at the axial slice of the L1 vertebra on CT images (L1-vBMD). Then, the L1-L4 vertebrae were segmented from the CT images to measure the three-dimensional BMD at the trabecular region of the L1-L4 vertebral bodies (CT-vBMD). Lastly, the segmented vertebrae were projected onto the coronal plane to measure the two-dimensional BMD (CT-aBMD). Each parameter was correlated with DXA-BMD, and the receiver operating characteristic (ROC) curve to diagnose osteoporosis was assessed. RESULTS: The correlation coefficients of DXA-BMD with L1-vBMD, CT-vBMD, and CT-aBMD were 0.364, 0.456, and 0.911, respectively (all p < 0.01). In the ROC curve analysis to diagnose osteoporosis, the area under the curve for CT-aBMD (0.941) was significantly higher than those for L1-vBMD (0.582) and CT-vBMD (0.657) (both p < 0.01). CONCLUSION: Compared with L1-vBMD and CT-vBMD, CT-aBMD could accurately predict DXA-BMD and detect patients with osteoporosis. Given that our method can quantify BMD in both two and three dimensions, it could be used to screen for osteoporosis from quantitative CT images.