Artificial intelligence ChatGPT in ophthalmology and scientific writing / 国际眼科杂志(Guoji Yanke Zazhi)

ChatGPT is a large language models(LLMs)that uses deep learning techniques to produce human-like responses to natural language inputs. It belongs to the family of generative pre-training transformer(GPT)models currently publicly available developed by OpenAI in November 2022. ChatGPT is capable of capturing the nuances and intricacies of human language, generating appropriate and contextually relevant responses. It can assist medical professionals in various tasks, such as research, diagnosis, patient monitoring, and medical education, from identifying research programs to assisting in clinical and laboratory diagnosis, to know new developments in their fields and scientific writing. ChatGPT has also attracted increasing attention and widely used in ophthalmology. However, the use of ChatGPT and other artificial intelligence tools in such tasks comes now with several limitations, ethical and legal concerns, such as credibility, plagiarism, copyright infringement, and biases. Future research can focus on developing new methods to mitigate these limitations while harnessing the benefits of ChatGPT in medicine and related aspects.

Research and application status of intelligent auxiliary clinical diagnosis and treatment equipment / 中国医学装备

Artificial intelligence(AI)is a strategic technology leading a new round of technological revolution and industrial transformation.It is forward-looking,important and necessary to apply AI technology to the medical field.At present,the research and development of intelligent data monitoring equipment,intelligent medical instruments,disease auxiliary diagnosis and treatment platforms,auxiliary diagnosis and treatment integrated systems and other technologies have been widely carried out,and related products are gradually used in auxiliary medical prevention,diagnosis,treatment,and rehabilitation.Based on the recent development of AI technology in the medical field,the application status of intelligent auxiliary diagnosis and treatment equipment in three fields of intelligent monitoring equipment,virtual psychological diagnosis and treatment platforms,and traditional Chinese medicine auxiliary diagnosis and treatment instruments was summarized to provide reference for AI to connect disease and health,and realize the intersection of AI and medical disciplines.

Study on the effects of AI reconstruction algorithm based on different tube voltages on CT imaging quality and radiation dose of chest phantom / 中国医学装备

Objective:To investigate the effects of different tube voltages combined with artificial intelligence reconstruction algorithm(CI)on the computed tomography(CT)imaging quality and radiation dose of chest phantom on the basis of the CT scan for an adult male simulated chest phantom(PH-N1).Methods:A 512-slice CT scanner of ultrahigh-end was adopted to conduct scan,and the images were divided into 70 kV group,80 kV group,100 kV group and 120 kV group according to different tube voltage.For 4 groups of CT scan images with different tube voltages,the 10%,30%,50%,70%and 90%CI were adopted to reconstruct 1mm thin layer image.The CT volume dose index(CTDIvol)and the dose-length product(DLP)of the scans of 4 groups were compared.The CT values and standard deviation(SD)values of the aorta,abdominal wall fat and erector spine muscle were measured.Two senior diagnostic physicians with more than 5 years of work experience independently and double-blindly evaluated the image quality by using 5-point scale.A Kappa consistency test was conducted.One-way analysis of variance was adopted to compare the differences of CT values and SD values of the tissues of image targets.The Friedman rank-sum test was adopted to compare the differences of subjective image qualities among different groups.Results:The differences of CTDIvol and DLP among 4 groups with different tube voltages were significant(F=1855.617,3996.118,P<0.05),respectively.Under 70 kV tube voltage,there were no significant differences in CT values of the aorta,abdominal wall fat and erector spine muscle,which were reconstructed by using 10%,30%,50%,70%and 90%CI(P>0.05),while the differences of SD values among them were statistically significant(F=32.267,53.327,14.873,P<0.05),respectively.Under the different tube voltages of 4 groups,which were reconstructed by 90%CI,the CT values of aorta,abdominal wall fat and erector spine muscle gradually decreased with decreasing of tube voltage,the differences were significant(F=139.899,2563.93,219.231,P<0.05),respectively.The consistency of subjective scores between two diagnostic physicians was better for each group of images(Kappa=0.712～0.869).Conclusion:Compared with 80 kV,90 kV and 120 kV images,the reconstructed images with 90%CI algorithm under 70 kV tube voltage can significantly reduce the radiation dose,and the images have a favorable signal-to-noise ratio at the same time.

Effectiveness of the artificial intelligence image recognition system in diagnosing endometrial cytopathology / 西安交通大学学报(医学版)

【Objective】 To explore the effectiveness of an image recognition system based on artificial intelligence (AI) in diagnosing benign and malignant endometrial cell clumps. 【Methods】 We selected endometrial cytological specimens from The First Affiliated Hospital of Xi’an Jiaotong University and Xi’an Daxing Hospital from August 2021 to February 2023; histopathology was used as the gold standard. We compared and analyzed the sensitivity, specificity, positive predictive value, negative predictive value, accuracy and diagnostic time of AI image recognition system (AI diagnosis) and professional pathologists’ manual diagnosis (manual diagnosis) of benign and malignant endometrial cell clumps. 【Results】 Among the 126 patients included in the analysis, the overall coincidence rate of AI diagnosis and histological diagnosis was 92.1% (116/126), which was highly consistent with histopathological results (Kappa=0.841). The overall coincidence rate of manual diagnosis and histological diagnosis was 94.4% (119/126), which was highly consistent with histopathological results (Kappa=0.889). There was no statistically significant difference between AI diagnosis and manual diagnosis methods (χ2=0.568, P=0.451). The sensitivity, specificity, positive predictive value, and negative predictive value of AI diagnosis were 91.8%, 92.3%, 91.8%, and 92.3%, respectively. There were 126 cytology sections, each of which required 6.67 minutes for manual diagnosis and 5.00 minutes for AI diagnosis. 【Conclusion】 The AI image recognition system has high diagnostic accuracy, sensitivity and specificity, which is equivalent to the manual diagnosis level of professional pathologists. Therefore, this system has application value in the diagnosis of benign and malignant endometrial cell clumps.

Application status of artificial intelligence techniques in retinoblastoma / 国际眼科杂志(Guoji Yanke Zazhi)

Retinoblastoma is a kind of malignant eye tumor commonly seen in children, which is one of the main causes threatening children's vision and life. The diagnosis and evaluation of retinoblastoma has always been a hot topic in clinic. In the past few years, the application of artificial intelligence(AI)technology has made significant progress in the medical field, providing new opportunities and challenges for the diagnosis and treatment of retinoblastoma, for example, the use of AI algorithms to analyze massive clinical data, which can help doctors diagnose the disease more accurately and provide personalized treatment plans. In addition, AI technology also plays an important role in medical image analysis, genomics research and other aspects, which can help the development of new drugs and improve patient prognosis. This article reviews the application progress of AI in retinoblastoma.

Application of deep learning artificial intelligence in the auxiliary diagnosis of age-related macular degeneration / 国际眼科杂志(Guoji Yanke Zazhi)

Since the advent of artificial intelligence(AI), it has been increasingly applied and rapidly developed in various fields. In the field of medicine, image features can be automatically extracted and the performance of feature learning and classification can be completed with the help of AI. In the field of ocular fundus disease, AI can give a diagnosis of age-related maculopathy by analyzing and identifying fundus photography and optical coherence tomography with an accuracy rate similar to that of ophthalmologists. In the future, AI may assist physicians in making a diagnosis of age-related macular degeneration, aid basic hospital in screening and curb its progression in the early stage of the disease. However, the technique has problems such as uncertain model recognition performance, opaque operation process, and excessive amount of clinical data required, which still cannot be widely used in the clinic. In recent years, a lot of research has been done in China in the application of deep learning with AI to assist diagnosis of ophthalmic diseases, and the results show that AI combined with imaging analysis of ophthalmic diseases has such characteristics as objectivity, efficiency and accuracy. In this article, studies on deep learning in the auxiliary diagnosis of age-related maculopathy are reviewed, and the progress on its application and the limitations that exist are analyzed, so as to provide more information on the use and extension of AI in this disease.

Construction and Application of Rib Fracture Diagnosis Model Based on YOLOv3 Algorithm / 法医学杂志

OBJECTIVES@#The artificial intelligence-aided diagnosis model of rib fractures based on YOLOv3 algorithm was established and applied to practical case to explore the application advantages in rib fracture cases in forensic medicine.@*METHODS@#DICOM format CT images of 884 cases with rib fractures caused by thoracic trauma were collected, and 801 of them were used as training and validation sets. A rib fracture diagnosis model based on YOLOv3 algorithm and Darknet53 as the backbone network was built. After the model was established, 83 cases were taken as the test set, and the precision rate, recall rate, F1-score and radiology interpretation time were calculated. The model was used to diagnose a practical case and compared with manual diagnosis.@*RESULTS@#The established model was used to test 83 cases, the fracture precision rate of this model was 90.5%, the recall rate was 75.4%, F1-score was 0.82, the radiology interpretation time was 4.4 images per second and the identification time of each patient's data was 21 s, much faster than manual diagnosis. The recognition results of the model was consistent with that of the manual diagnosis.@*CONCLUSIONS@#The rib fracture diagnosis model in practical case based on YOLOv3 algorithm can quickly and accurately identify fractures, and the model is easy to operate. It can be used as an auxiliary diagnostic technique in forensic clinical identification.

Exploration and analysis on the influence of different image algorithms of chest CT on the detection effect of AI-assisted diagnostic software for pulmonary nodules / 中国医学装备

Objective:To explore the influence of artificial intelligence(AI)assisted diagnosis system on the detection effect of computed tomography(CT)on chest for pulmonary nodules under different image algorithms.Methods:The images of 100 patients who underwent CT scan of chest in hospital during 2020 and 2022 were analyzed.The original data were reconstructed by standard algorithm,lung algorithm and bone algorithm,and the AI-assisted diagnosis system was used to identify the data of three algorithms.The image data were transmitted to the AI-assisted diagnosis system,and then,the comprehensive judgment results of three physicians with above the title of attending physician were used as the"gold standard"to compare the detection effect and nodule nature of AI.The detection rates and false detection rates of pulmonary nodules were calculated under different image algorithms and the combination of different algorithm.Results:There was a statistically significant difference in the detection rates of solid pulmonary nodule and partial solid pulmonary nodule under different image algorithms(F=262.64,F=440.74,P<0.05),in which the simultaneous recognition of lung algorithm and bone algorithm had the highest detection rate.There was no statistically significant difference in the false detection rates of solid pulmonary nodule and partial solid pulmonary nodule among different algorithms,and there was no statistically significant difference in measured diameters of solid pulmonary nodules and partial solid pulmonary nodule among different algorithms,and there were statistically significant differences in CT values of pulmonary nodule among three image algorithms(F=82.42,F=63.08,P<0.05).Conclusion:Compared with other algorithm,the Bone algorithm of CT scan on chest has more detective effect for the detections of solid and partial solid pulmonary nodule(include ground glass nodule),and the qualities of complex pulmonary nodules.But the overall detection effect and false detection rate of that are lower than those of Lung algorithm.For high-risk groups,the Bone algorithm or high-resolution algorithm can be added during undergoing CT examination on chest,which can implement artificial intelligence(AI)systematic recognition with Lung algorithm,and optimize the detection effect of AI system.

Evaluation and impact analysis of ChatGPT′s competence in ethical review management / 中华医学科研管理杂志

Objective:This paper tried to have a dialogue with ChatGPT about ethics review, to understand the degree of intelligence of this application in the field of ethical management, and to analyze its possible impact on the future ethics review work.Methods:The research team sorted out 43 questions, then the research team member at abroad accomplish the dialogue with ChatGPT in both Chinese and English. Feedback answers were summarized and analyzed to explore their advantages and problems.Results:Most of the ChatGPT′s answers of this test were reasonable, with obvious advantages in response speed, and the rigor and friendliness were relatively good. However, there were also problems in consistency, comprehensiveness and expertise, the accuracy and computing power also still have a lot of space for improvement.Conclusions:It is too early for AI to replace professionals, but we should fully develop and utilize the advantages of AI to help professionals get rid of inefficient labor and play a better role.

Development and accuracy of artificial intelligence-generated prediction of facial changes in orthodontic treatment: a scoping review / 浙江大学学报（英文版）（B辑：生物医学和生物技术）

Artificial intelligence (AI) has been utilized in soft-tissue analysis and prediction in orthodontic treatment planning, although its reliability has not been systematically assessed. This scoping review was conducted to outline the development of AI in terms of predicting soft-tissue changes after orthodontic treatment, as well as to comprehensively evaluate its prediction accuracy. Six electronic databases (PubMed, EBSCOhost, Web of Science, Embase, Cochrane Library, and Scopus) were searched up to March 14, 2023. Clinical studies investigating the performance of AI-based systems in predicting post-orthodontic soft-tissue alterations were included. The Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) and Joanna Briggs Institute (JBI) appraisal checklist for diagnostic test accuracy studies were applied to assess risk of bias, while the Grading of Recommendation, Assessment, Development, and Evaluation (GRADE) assessment was conducted to evaluate the certainty of outcomes. After screening 2500 studies, four non-randomized clinical trials were finally included for full-text evaluation. We found a low level of evidence indicating an estimated high overall accuracy of AI-generated prediction, whereas the lower lip and chin seemed to be the least predictable regions. Furthermore, the facial morphology simulated by AI via the fusion of multimodality images was considered to be reasonably true. Since all of the included studies that were not randomized clinical trials (non-RCTs) showed a moderate to high risk of bias, more well-designed clinical trials with sufficient sample size are needed in future work.

Research on the construction and application value of artificial intelligent recognition model of nasal fracture / 中国法医学杂志

Objective The diagnosis of nasal fractures poses challenges in forensic clinical evaluation.This study aims to develop and enhance an artificial intelligence-based model for nasal fracture recognition,evaluate its performance,and provide assistance and support for forensic clinical identification.Methods Multi-center nasal CT images were selected and screened according to the consensus standards set by Chinese experts in nasal CT examination and diagnosis.A recognition model was constructed,followed by external verification and evaluation.Additionally,the diagnostic capabilities of qualified appraisers/doctors with different professional titles(primary,intermediate,and senior)were compared with the performance of the intelligent recognition model.The accuracy,sensitivity,specificity),and negative predictive value(NP)of the intelligent recognition model were comprehensively evaluated.Results The intelligent recognition model exhibited high diagnostic efficiency and stability.It improved the diagnostic accuracy of radiologists and appraisers in detecting nasal fractures while effectively bridging the gap between inexperienced doctors/appraisers and experienced ones.Conclusion The intelligent recognition model for nasal fractures can assist appraisers in enhancing their ability to locate such fractures on CT images and improve work efficiency while enhancing appraisal opinions'accuracy and scientificity.

Ideologia gerencialista e plataformas de treinamentos de dados para Inteligência Artificial (IA): condições de trabalho e saúde dos trabalhadores no Brasil / Managerial ideology and Artificial Intelligence (AI) data training platform: working conditions and worker's health in Brazil / Ideología managerial y plataformas de anotación de datos para Inteligencia Artificial (IA): condiciones de trabajo y salud de los trabajadores en Brasi

O objetivo deste estudo é analisar as condições de trabalho e os seus impactos na saúde dos trabalhadores no mercado de microtarefas de treinamento de dados para a produção de Inteligência Artificial (IA), em especial no que diz respeito a suas relações com a ideologia gerencialista. Os dados são provenientes de uma netnografia realizada entre os anos de 2020 e 2021, de análises dos websites das plataformas e de entrevistas realizadas com 15 trabalhadores. A partir da análise de quatro instâncias mediadoras (econômica, política, ideológica e psicológica), argumentamos que a ideologia gerencialista, consubstanciada a ideologia californiana, se caracteriza como um operador central na gestão do trabalho, que tem por finalidade garantir a adesão dos trabalhadores às plataformas e ocultar os conflitos do trabalho, direcionando-os para o nível individual e produzindo um cenário de individualização do sofrimento.

The objective of this study is to analyze working conditions and their impacts on worker's health in the Artificial Intelligence (AI) data annotation microtask market, especially to highlight their relationship with managerial ideology. The data comes from a netnography carried out between the years 2020 and 2021, from analysis on the platform's websites, and from interviews with 15 workers. Drawing from the analysis of four different mediation systems (economic, political, ideological, and psychological), we argue that the managerial ideology, overlaid with the Californian ideology, is characterized as a central element in the management of labor, which aims to guarantee the adherence of workers to platforms and hide the labor conflicts, directing them to the individual level and producing a scenario of individualization of suffering.

El objetivo de esta investigación es analizar las condiciones de trabajo y sus impactos en la salud de los tra-bajadores en el mercado de microtareas de anotación de datos para la producción de Inteligencia Artificial (IA), en particular en lo que concierne a su relación con la ideología managerial. Los datos provienen de una netnografía realizada entre los años 2020 y 2021, de análisis en los sitios web de las plataformas y de entrevistas con 15 trabajadores. A partir del análisis de cuatro instancias mediadoras (económica, política, ideológica y psicológica), argumentamos que la ideología gerencial, superpuesta en la ideología californi-ana, se caracteriza como un elemento central en la gestión del trabajo, que pretende garantizar la adhesión de los trabajadores a las plataformas y ocultar los conflictos del trabajo, dirigiéndolos al plano individual y produciendo un escenario de individualización del sufrimiento.

Artificial intelligence and its application for cardiovascular diseases in Chinese medicine / 数字中医药（英文）

@#Cardiovascular diseases (CVDs) are major disease burdens with high mortality worldwide. Early prediction of cardiovascular events can reduce the incidence of acute myocardial infarction and decrease the mortality rates of patients with CVDs. The pathological mechanisms and multiple factors involved in CVDs are complex; thus, traditional data analysis is insufficient and inefficient to manage multidimensional data for the risk prediction of CVDs and heart attacks, medical image interpretations, therapeutic decision-making, and disease prognosis prediction. Meanwhile, traditional Chinese medicine (TCM) has been widely used for treating CVDs. TCM offers unique theoretical and practical applications in the diagnosis and treatment of CVDs. Big data have been generated to investigate the scientific basis of TCM diagnostic methods. TCM formulae contain multiple herbal items. Elucidating the complicated interactions between the active compounds and network modulations requires advanced data-analysis capability. Recent progress in artificial intelligence (AI) technology has allowed these challenges to be resolved, which significantly facilitates the development of integrative diagnostic and therapeutic strategies for CVDs and the understanding of the therapeutic principles of TCM formulae. Herein, we briefly introduce the basic concept and current progress of AI and machine learning (ML) technology, and summarize the applications of advanced AI and ML for the diagnosis and treatment of CVDs. Furthermore, we review the progress of AI and ML technology for investigating the scientific basis of TCM diagnosis and treatment for CVDs. We expect the application of AI and ML technology to promote synergy between western medicine and TCM, which can then boost the development of integrative medicine for the diagnosis and treatment of CVDs.

Artificial intelligence in thoracic surgery / 中国胸心血管外科临床杂志

@#As an emerging technology, artificial intelligence (AI) uses human theory and technology for robots to study, develop, learn and identify human technologies. Thoracic surgeons should be aware of new opportunities that may affect their daily practice by the direct use of AI technology, or indirect use in the relevant medical fields (radiology, pathology, and respiratory medicine). The purpose of this paper is to review the application status and future development of AI associated with thoracic surgery, diagnosis of AI-related lung cancer, prognosis-assisted decision-making programs and robotic surgery. While AI technology has made rapid progress in many areas, the medical industry only accounts for a small part of AI use, and AI technology is gradually becoming widespread in the diagnosis, treatment, rehabilitation, and care of diseases. The future of AI is bright and full of innovative perspectives. The field of thoracic surgery has conducted valuable exploration and practice on AI, and will receive more and more influence and promotion from AI.

Clinical application of artificial intelligence to lung nodules diagnosis in regional medical center / 中国胸心血管外科临床杂志

@#Objective    To explore the efficacy of artificial intelligence (AI) detection on pulmonary nodule compared with multidisciplinary team (MDT) in regional medical center. Methods    We retrospectively analyzed the clinical data of 102 patients with lung nodules in the Xiamen Fifth Hospital from April to December 2020. There were 57 males and 45 females at age of 36-90 (48.8±11.6) years. The preoperative chest CT was imported into AI system to record the detected lung nodules. The detection rate of pulmonary nodules by AI system was calculated, and the sensitivity, specificity of AI in the different diagnosis of benign and malignant pulmonary was calculated and compared with manual film reading by MDT. Results    A total of 322 nodules were detected by AI software system, and 305 nodules were manually detected by physicians (P<0.05). Among them, 113 pulmonary nodules were diagnosed by pathologist. Thirty-eight of 40 lung cancer nodules were AI high-risk nodules, the sensitivity was 95.0%, and 25 of 73 benign nodules were AI high-risk nodules, the specificity was 65.8%. Lung cancer nodules were correctly diagnosed by MDT, but  benign nodules were still considered as  lung cancer at the first diagnosis in 10 patients. Conclusion    AI assisted diagnosis system has strong performance in the detection of pulmonary nodules, but it can not content itself with clinical needs in the differentiation of benign and malignant pulmonary nodules. The artificial intelligence system can be used as an auxiliary tool for MDT to detect pulmonary nodules in regional medical center.

Clinically applicable artificial intelligence algorithm for the diagnosis, evaluation, and monitoring of acute retinal necrosis / 浙江大学学报（英文版）（B辑：生物医学和生物技术）

The prompt detection and proper evaluation of necrotic retinal region are especially important for the diagnosis and treatment of acute retinal necrosis (ARN). The potential application of artificial intelligence (AI) algorithms in these areas of clinical research has not been reported previously. The present study aims to create a computational algorithm for the automated detection and evaluation of retinal necrosis from retinal fundus photographs. A total of 149 wide-angle fundus photographs from 40 eyes of 32 ARN patients were collected, and the U-Net method was used to construct the AI algorithm. Thereby, a novel algorithm based on deep machine learning in detection and evaluation of retinal necrosis was constructed for the first time. This algorithm had an area under the receiver operating curve of 0.92, with 86% sensitivity and 88% specificity in the detection of retinal necrosis. For the purpose of retinal necrosis evaluation, necrotic areas calculated by the AI algorithm were significantly positively correlated with viral load in aqueous humor samples (

Application of artificial intelligence in lung tumor imaging diagnosis / 中国肿瘤临床

Early detection and accurate diagnosis are critical for the prognosis of lung cancer. Radiological imaging could reflect tumor heterogeneity in a non-invasive and comprehensive manner. Deep mining of high throughput imaging data is a big challenge for radiologists. Artificial intelligence (AI) methods excel at processing large quantities of high-dimensional information and analyzing data using algorithm. It can automatically recognize complex patterns in imaging data, provide quantitative assessments of radiographic characteristics, and is promising in tumor detection and diagnosis. Precision medicine could be made when AI was integrated into the clinical workflow as a tool to assist radiologists. Here we review the current progress and discuss the challenges and future directions of AI applications in lung tumor imaging diagnosis.

Discussion on value of medical records-structured specialized disease database based on artificial intelligence in clinical research / 上海交通大学学报（医学版）

Objective: To explore the value support of medical records-structured specialized disease database established by using unstructured electronic medical record text information in clinical research. Methods: The information of patients who were admitted to a Grade A specialist hospital in Shanghai from Oct. 2007 to Sept. 2019 were collected. By using artificial intelligence (AI) engine and other information methods, the electronic medical record text information were structured into a structured database, and compared with the traditional structured database. Results: The information of 82 584 patients were collected, and the structured number of hospital records was 253 000. The specialized disease databases of lung cancer, esophageal cancer and mediastinal tumor were established. Compared with the traditional structured database, the specialized disease database expanded the scope of data retrieval and improved the efficiency of data retrieval. Conclusion: The construction of medical records-structured specialized disease database based on AI reduces the burden of clinician data retrieval, and provides valuable statistical data for clinical research.