Objective The high rates of missed diagnosis and misdiagnosis limit the diagnosis of femoral neck fracture (FNF), which requires a new method to assist doctors to get more accurate diagnosis of FNF. This study aims to estimate the ability of AI in the detection of FNF and further compare its performance with human level. And the performance of AI-aided human level is also explored to confirm the value of AI as an assistant for clinical doctors to detect the FNF. Materials and Methods 4,477 hip X-rays (consisted of 2,884 FNF X-rays and 1,593 normal hip X-rays) from eight Chinese top tree hospitals (Union Hospital, Tongji Medical College, Huazhong University of Science and Technology (Wuhan Union Hospital), Wuhan Pu’ai Hospital, Tianyou Hospital, Wuhan University of Science and Technology, Hanyang Hospital, Wuhan University of Science and Technology, Northern Jiangsu People’s Hospital, Xiangya Changde Hospital, People's Hospital of Tibet Autonomous Region and the Second Affiliated Hospital of Soochow University) were collected to establish a large multi-center clinical sample database. Then the X-rays were labeled, and the database was divided into training dataset (4,029 X-rays) and testing dataset (448 X-rays). A Faster RCNN model with three different backbones (VGG16, VGG16-nottop and Resnet 50) was set up and trained with the training dataset, then the diagnostic performance of the Faster RCNN was assessed by the testing dataset and further compared with five doctors, in the form of accuracy, sensitivity, specificity, missed diagnosis rate, misdiagnosis rate, positive predictive value (PPV), negative predictive value (NPV), and time consumption. The result of the backbone with best performance was further set as reference for the doctor to diagnose the testing dataset again to confirm the value of AI as an assistant to detect the FNF. Results Faster RCNN with Resnet 50 performed best compared with the other two backbones (VGG16 performed lowest, VGG16-nottop performed at medium level) in accuracy (0.82 vs 0.58 and 0.76), sensitivity (0.93 vs 0.83 and 0.94), specificity (0.62 vs 0.12 and 0.43), missed diagnosis rate (0.07 vs 0.17 and 0.06), misdiagnosis rate (0.38 vs 0.88 and 0.57), PPV (0.82 vs 0.63 and 0.75), NPV (0.82 vs 0.28 and 0.81) and time consumption (0.02h vs 0.04 h and 0.03h). And compared with human level, the Faster RCNN with Resnet 50 expressed better ability in terms of accuracy, sensitivity, missed diagnosis rate, NPV and time consumption, and worse ability in specificity and misdiagnosis rate. As for the PPV, there was not significant difference. Under the assistance of Faster RCNN with Resnet 50, the human level was enhanced in all aspects. Conclusion As a new application of intelligent medicine, AI can be qualified in the detection of FNF, and can also be an excellent assistant for clinical doctors to improve the diagnosis of FNF.

OBJECTIVE:This study aimed to evaluate and compare the effectiveness of knowledge base-optimized and unoptimized large language models (LLMs) in the field of orthopedics to explore optimization strategies for the application of LLMs in specific fields. METHODS:This research constructed a specialized knowledge base using clinical guidelines from the American Academy of Orthopaedic Surgeons (AAOS) and authoritative orthopedic publications. A total of 30 orthopedic-related questions covering aspects such as anatomical knowledge, disease diagnosis, fracture classification, treatment options, and surgical techniques were input into both the knowledge base-optimized and unoptimized versions of the GPT-4, ChatGLM, and Spark LLM, with their generated responses recorded. The overall quality, accuracy, and comprehensiveness of these responses were evaluated by 3 experienced orthopedic surgeons. RESULTS:Compared with their unoptimized LLMs, the optimized version of GPT-4 showed improvements of 15.3% in overall quality, 12.5% in accuracy, and 12.8% in comprehensiveness; ChatGLM showed improvements of 24.8%, 16.1%, and 19.6%, respectively; and Spark LLM showed improvements of 6.5%, 14.5%, and 24.7%, respectively. CONCLUSION:The optimization of knowledge bases significantly enhances the quality, accuracy, and comprehensiveness of the responses provided by the 3 models in the orthopedic field. Therefore, knowledge base optimization is an effective method for improving the performance of LLMs in specific fields.

OBJECTIVE:To evaluate the accuracy and parsing ability of GPT 4.0 for Japanese medical practitioner qualification examinations in a multidimensional way to investigate its response accuracy and comprehensiveness to medical knowledge. METHODS:We evaluated the performance of the GPT 4.0 on Japanese Medical Licensing Examination (JMLE) questions (2021-2023). Questions are categorized by difficulty and type, with distinctions between general and clinical parts, as well as between single-choice (MCQ1) and multiple-choice (MCQ2) questions. Difficulty levels were determined on the basis of correct rates provided by the JMLE Preparatory School. The accuracy and quality of the GPT 4.0 responses were analyzed via an improved Global Qualily Scale (GQS) scores, considering both the chosen options and the accompanying analysis. Descriptive statistics and Pearson Chi-square tests were used to examine performance across exam years, question difficulty, type, and choice. GPT 4.0 ability was evaluated via the GQS, with comparisons made via the Mann-Whitney U or Kruskal-Wallis test. RESULTS:The correct response rate and parsing ability of the GPT4.0 to the JMLE questions reached the qualification level (80.4%). In terms of the accuracy of the GPT4.0 response to the JMLE, we found significant differences in accuracy across both difficulty levels and option types. According to the GQS scores for the GPT 4.0 responses to all the JMLE questions, the performance of the questionnaire varied according to year and choice type. CONCLUSION:GTP4.0 performs well in providing basic support in medical education and medical research, but it also needs to input a large amount of medical-related data to train its model and improve the accuracy of its medical knowledge output. Further integration of ChatGPT with the medical field could open new opportunities for medicine.

Artificial intelligence (AI) is an interdisciplinary field that combines computer technology, mathematics, and several other fields. Recently, with the rapid development of machine learning (ML) and deep learning (DL), significant progress has been made in the field of AI. As one of the fastest-growing branches, DL can effectively extract features from big data and optimize the performance of various tasks. Moreover, with advancements in digital imaging technology, DL has become a key tool for processing high-dimensional medical image data and conducting medical image analysis in clinical applications. With the development of this technology, the diagnosis of orthopedic diseases has undergone significant changes. In this review, we describe recent research progress on DL in fracture diagnosis and discuss the value of DL in this field, providing a reference for better integration and development of DL technology in orthopedics.

When it comes to object detection tasks, YOLO stands out for its impressive speed and efficiency. Nonetheless, deploying YOLO on resource-constrained devices remains a challenge due to its substantial model size and memory requirements. The direct application of conventional binary quantization strategies to YOLO can result in significant accuracy degradation. A prevalent solution is to introduce floating-point shortcuts. However, the increased computational demand and parameter complexity associated with these shortcuts limit their practical deployment on hardware platforms for optimal acceleration. To solve this problem, we propose a binary neural network (BNN) for object detection called BOB-YOLO to achieve a balanced performance in terms of computational speed, model size, and detection accuracy. Our BOB-YOLO fully leverages module-wise latency (MWL) to supervise the latency of floating-point shortcut branches by that of 1-bit trunk branches. This supervision maximizes the information carried by the floating-point data flow in shortcuts while maintaining latency within the limits set by the 1-bit convolution branch, thereby improving parallel computational efficiency. We also introduce the Roofline Model to address these limitations by considering both computational complexity and parameter compression, ensuring high computational intensity. Additionally, we propose a performance evaluation metric Pd, which provides an intuitive description of the trade-off between speed and accuracy, aligning closely with the practical requirements of binary quantization strategies. Extensive experiments on the VOC and COCO datasets demonstrate the significant advantages of our method over state-of-the-art BNN methods.

The medical metaverse is a combination of medicine, computer science, information technology and other cutting-edge technologies. It redefines the method of information interaction about doctor-patient communication, medical education and research through the integration of medical data, knowledge and services in a virtual environment. Artificial intelligence (AI) is a discipline that uses computer technology to study and develop human intelligence. AI has infiltrated every aspect of medical metaverse and is deeply integrated with the technologies that build medical metaverse, such as large language models (LLMs), digital twins, blockchain and extended reality (including VR/AR/XR). AI has become an integral part of the medical metaverse building process. Moreover, AI also provides richer medical metaverse functions, including diagnosis, education, and consulting. This paper aims to introduce how AI supports the development of medical metaverse, including its specific application scenarios, shortcomings and future development. Our goal is to contribute to the advancement of more sophisticated and intelligent medical methods.

Millimeter waves are electromagnetic waves with wavelengths of 1–10 mm, which have characteristics of high frequency and short wavelength. They have gradually and widely been used in engineering and medical fields. We have identified studies related to millimeter waves in the biomedical field and summarized the biological effects of millimeter waves and their current status in medical applications. Finally, the shortcomings of existing studies and future developments were analyzed and discussed, with the aim of providing a reference for further research and development of millimeter waves in the medical field.

OBJECTIVE:This study aims to evaluate the instructional efficacy of a 3D Surgical Training System (3DSTS), which combines real surgical footage with high-definition 3D animations, against conventional surgical videos and textbooks in the context of orthopedic proximal humerus fracture surgeries. DESIGN:Before the experiment, 89 participants completed a pre-educational knowledge assessment. They were then randomized into 3 groups: the 3DSTS group (n = 30), the surgical video (SV) group (n = 29), and the textbook group (n = 30). After their respective teaching courses, all participants took a posteducational assessment and completed a perceived cognitive load test. The 3DSTS group also filled out a satisfaction survey. Once all assessments were finished, the SV and textbook groups were introduced to the 3DSTS course and subsequently completed a satisfaction survey. All statistical analyses were executed using IBM SPSS version 24 (IBM Corp., Armonk, NY). For data fitting normal distribution, we employed one-way analysis of variance (one-way ANOVA) and Tukey HSD tests, whereas, for non-normally distributed data, we used Kruskal-Wallis H tests and Dunn's tests. The significance level for all tests was set at p < 0.05. SETTING:Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei Province, P. R. China. PARTICIPANTS:About 89 doctors who undergoing standardized residents training. RESULT:The initial assessment scores among the three groups were comparable, showing no significant statistical difference. Post-education revealed a marked difference in the scores, with the 3DSTS group outperforming both the SV and textbook groups. Specifically, the 3DSTS group exhibited statistically greater improvement in areas such as procedural steps, and specialized surgical techniques compared to the SV and textbook groups. During the 3DSTS teaching process, participants reported the least perceived cognitive load and expressed strong satisfaction, highlighting that the instructional materials are well-prepared, and considering this teaching method superior and more innovative than previous courses they had encountered. CONCLUSION:The 3D Surgical Training System, integrating real videos with 3D animations, significantly enhances orthopedic surgery education over conventional methods, providing improved comprehension, lower cognitive load, and standardized learning outcomes. Its efficacy and high participant satisfaction underscore its potential for broader adoption in surgical disciplines. This study is registered with ClinicalTrials. gov ID: ChiCTR2300074730.

You Only Look Once (YOLO), known for its real-time performance and outstanding accuracy, has emerged as a prominent framework for object detection tasks. However, deploying YOLO on resource-constrained edge devices poses challenges due to its substantial memory requirements. In this paper, we propose MPQ-YOLO, an ultra-low mixed-precision quantization framework designed for edge device deployment. The core idea is to integrate 1-bit Backbone quantization and 4-bit Head quantization with dedicated training techniques. Specifically, we analyze the effect of numerical distribution on the performance of binary neural networks (BNNs), and based on this, we design a backbone with only 1-bit convolution. Then, we introduce a trainable scale and Progressive Network Quantization (PNQ) training strategy to bridge the Backbone and Head for end-to-end quantization training. The former is applied to both weights and activations within the 4-bit Head, enabling effective gradient propagation. The latter mitigates oscillation caused by mixed precision training, promoting smoother training and faster model convergence. Extensive experiments on VOC and COCO datasets demonstrate that MPQ-YOLO achieves a good trade-off between model compression and detection performance. Specifically, compared to the full-precision model, MPQ-YOLO achieves compression of up to 16.3× and 14.2× in terms of computational complexity and model size, respectively, while maintaining relatively high detection accuracy, i.e., 74.7% on VOC and 51.5% on COCO. To the best of our knowledge, MPQ-YOLO is the first YOLO framework with dual low mixed-precision quantization. Moreover, compared to the existing layer-wise mixed-precision quantization methods which cause redundant data processing and massive data movement, MPQ-YOLO offers a more hardware-designer-friendly and straightforward solution through efficient resource utilization and reuse.

Objectives:To explore an intelligent detection technology based on deep learning algorithms to assist the clinical diagnosis of distal radius fractures (DRFs), and further compare it with human performance to verify the feasibility of this method. Methods:A total of 3,240 patients (fracture: n = 1,620, normal: n = 1,620) were included in this study, with a total of 3,276 wrist joint anteroposterior (AP) X-ray films (1,639 fractured, 1,637 normal) and 3,260 wrist joint lateral X-ray films (1,623 fractured, 1,637 normal). We divided the patients into training set, validation set and test set in a ratio of 7:1.5:1.5. The deep learning models were developed using the data from the training and validation sets, and then their effectiveness were evaluated using the data from the test set. Evaluate the diagnostic performance of deep learning models using receiver operating characteristic (ROC) curves and area under the curve (AUC), accuracy, sensitivity, and specificity, and compare them with medical professionals. Results:The deep learning ensemble model had excellent accuracy (97.03%), sensitivity (95.70%), and specificity (98.37%) in detecting DRFs. Among them, the accuracy of the AP view was 97.75%, the sensitivity 97.13%, and the specificity 98.37%; the accuracy of the lateral view was 96.32%, the sensitivity 94.26%, and the specificity 98.37%. When the wrist joint is counted, the accuracy was 97.55%, the sensitivity 98.36%, and the specificity 96.73%. In terms of these variables, the performance of the ensemble model is superior to that of both the orthopedic attending physician group and the radiology attending physician group. Conclusion:This deep learning ensemble model has excellent performance in detecting DRFs on plain X-ray films. Using this artificial intelligence model as a second expert to assist clinical diagnosis is expected to improve the accuracy of diagnosing DRFs and enhance clinical work efficiency.

As the most commonly used quantization techniques for deep neural networks, the int-only quantization methods use scale factor to linearly approximate the weights or activation of each layer. However, when passing activation data between layers, such int-only quantization methods require extra Scale Factor Conversion (SFC) operations, resulting in computational overhead. In this paper, we propose a Group-Wise Quantization framework, called GWQ, to reduce computational consumption during the activation data pass process by allowing multiple layers share one scale factor in SFC operations. Specifically, in the GWQ framework, we propose two algorithms for network layers grouping and model training. For the grouping of network layers, we propose a grouping algorithm based on the similarity of data numerical distribution. Then, the network layers divided into the same group will be quantified using the same common scale factor to reduce the computational consumption. Considering the additional performance loss caused by sharing scale factors among multiple layers, we propose a training algorithm to optimize these shared scale factors and model parameters, by designing a learnable power-of-two scaling parameter for each layer. Extensive experiments demonstrate that the proposed GWQ framework is able to effectively reduce the computational burden during inference, while maintaining model performance with negligible impact.

Objective: Explore a new deep learning (DL) object detection algorithm for clinical auxiliary diagnosis of lumbar spondylolisthesis and compare it with doctors’ evaluation to verify the effectiveness and feasibility of the DL algorithm in the diagnosis of lumbar spondylolisthesis.Methods: Lumbar lateral radiographs of 1,596 patients with lumbar spondylolisthesis from three medical institutions were collected, and senior orthopedic surgeons and radiologists jointly diagnosed and marked them to establish a database. These radiographs were randomly divided into a training set (n = 1,117), a validation set (n = 240), and a test set (n = 239) in a ratio of 0.7 : 0.15: 0.15. We trained two DL models for automatic detection of spondylolisthesis and evaluated their diagnostic performance by PR curves, areas under the curve, precision, recall, F1-score. Then we chose the model with better performance and compared its results with professionals’ evaluation.Results: A total of 1,780 annotations were marked for training (1,242), validation (263), and test (275). The Faster Region-based Convolutional Neural Network (R-CNN) showed better precision (0.935), recall (0.935), and F1-score (0.935) in the detection of spondylolisthesis, which outperformed the doctor group with precision (0.927), recall (0.892), f1-score (0.910). In addition, with the assistance of the DL model, the precision of the doctor group increased by 4.8%, the recall by 8.2%, the F1-score by 6.4%, and the average diagnosis time per plain X-ray was shortened by 7.139 s.Conclusion: The DL detection algorithm is an effective method for clinical diagnosis of lumbar spondylolisthesis. It can be used as an assistant expert to improve the accuracy of lumbar spondylolisthesis diagnosis and reduce the clinical workloads.

ABSTRACT In recent years, mixed reality (MR) technology has emerged as a new digital visualization tool and has been increasingly used in orthopedics. In comparison to traditional methods, MR provides a deeper integration of virtual information with the real world, offering users an intuitive display of anatomical details and facilitating the transition from two-dimensional to three-dimensional imaging data. With these significant features, MR has the potential to play a vital role in orthopedics. This article outlines the unique characteristics of MR and explores its technical advantages in doctor-patient communication, medical education, and surgical navigation within the field of orthopedics. This paper aims to inspire greater use of MR technology in the era of intelligent medicine.

As a high-tech strategic emerging comprehensive industry, the nuclear industry is committed to the research, production, and processing of nuclear fuel, as well as the development and utilization of nuclear energy. Nowadays, the nuclear industry has made remarkable progress in the application fields of nuclear weapons, nuclear power, nuclear medical treatment, radiation processing, and so on. With the development of artificial intelligence and the proposal of "Industry 4.0", more and more artificial intelligence technologies are introduced into the nuclear industry chain to improve production efficiency, reduce operation cost, improve operation safety, and realize risk avoidance. Meanwhile, deep learning, as an important technology of artificial intelligence, has made amazing progress in theoretical and applied research in the nuclear industry, which vigorously promotes the development of informatization, digitization, and intelligence of the nuclear industry. In this paper, we first simply comb and analyze the intelligent demand scenarios in the whole industrial chain of the nuclear industry. Then, we discuss the data types involved in the nuclear industry chain. After that, we investigate the research status of deep learning in the application fields corresponding to different data types in the nuclear industry. Finally, we discuss the limitation and unique challenges of deep learning in the nuclear industry and the future direction of the intelligent nuclear industry.

目的:探讨程序性死亡因子配体1(PD-L1)在骨肉瘤细胞和组织中的表达及对患者预后的影响.方法:采用Western Blot检测PD-L1在骨肉瘤细胞系和成骨细胞系中的表达,并用免疫组织化学(IHC)方法检测肿瘤及癌旁组织中PD-L1的表达,用Kaplan-Meier法进行生存分析.结果:PD-L1在骨肉瘤细胞株中的表达高于正常成骨细胞,而且其在骨肉瘤组织中的表达也显著高于癌旁组织,差异有统计学意义(P＜0.05).Kaplan-Meier生存分析显示PD-L1高表达患者预后较低表达患者差,差异有统计学意义(P＜0.05).结论:PD-L1在骨肉瘤中高表达,其高表达与患者预后不良有关,提示PD-L1可作为潜在的治疗靶点.

Chronic diseases are a growing concern worldwide, with nearly 25% of adults suffering from one or more chronic health conditions, thus placing a heavy burden on individuals, families, and healthcare systems. With the advent of the "Smart Healthcare" era, a series of cutting-edge technologies has brought new experiences to the management of chronic diseases. Among them, smart wearable technology not only helps people pursue a healthier lifestyle but also provides a continuous flow of healthcare data for disease diagnosis and treatment by actively recording physiological parameters and tracking the metabolic state. However, how to organize and analyze the data to achieve the ultimate goal of improving chronic disease management, in terms of quality of life, patient outcomes, and privacy protection, is an urgent issue that needs to be addressed. Artificial intelligence (AI) can provide intelligent suggestions by analyzing a patient's physiological data from wearable devices for the diagnosis and treatment of diseases. In addition, blockchain can improve healthcare services by authorizing decentralized data sharing, protecting the privacy of users, providing data empowerment, and ensuring the reliability of data management. Integrating AI, blockchain, and wearable technology could optimize the existing chronic disease management models, with a shift from a hospital-centered model to a patient-centered one. In this paper, we conceptually demonstrate a patient-centric technical framework based on AI, blockchain, and wearable technology and further explore the application of these integrated technologies in chronic disease management. Finally, the shortcomings of this new paradigm and future research directions are also discussed.

Real-time detection of sonar target plays a vital role in the underwater research field. Conventional deep learning methods need large quantities of sonar images as the sample for model training, and they cannot ensure detection speed and feature extraction ability simultaneously. For sonar dataset with small effective sample and low Signal-to-Noise Ratios (SNR), an improved YOLOv3 algorithm for real-time detection called as YOLOv3-DPFIN is proposed. The objective of the proposed YOLOv3-DPFIN is to accomplish the accurate detection of noise-intensive multi-category sonar targets with minimum time consumption. The proposed model conducts efficient feature extraction via the Dual-Path Network (DPN) module and the fusion transition module, and adopts a dense connection method to improve multi-scale prediction, which can complete precise object classification and location. The experimental results show that the algorithm achieves 84.4% ${mAP}_{75}$ with 56fps on a Nvidia Titan Xp, when testing on the sonar dataset and using the new VOC2012 mAP standard, which can meet the requirement of robust real-time detection for both raw and noised sonar targets. Moreover, the precision and speed of the proposed YOLOv3-DPFIN are superior to the original YOLOv3 model and state-of-the-art improved SSD models.