For the past few years, artificial intelligence (AI) technology has developed rapidly and has become frontier and hot topics in medical research. While the deep learning algorithm based on artificial neural networks is one of the most representative tool in this field. The advancement of ophthalmology is inseparable from a variety of imaging methods, and the pronounced convenience and high efficiency endow AI technology with promising applications in screening, diagnosis and follow-up of ophthalmic diseases. At present, related research on ophthalmologic AI technology has been carried out in terms of multiple diseases and multimodality. Many valuable results have been reported aiming at several common diseases of ophthalmology. It should be emphasized that ophthalmic AI products are still faced with some problems towards practical application. The regulatory mechanism and evaluation criteria have not yet integrated as a standardized system. There are still a number of aspects to be optimized before large-scale distribution in clinical utility. Briefly, the innovation of ophthalmologic AI technology is attributed to multidisciplinary cooperation, which is of great significance to China's public health undertakings, and will be bound to benefit patients in future clinical practice.
Objective To explore the clinical value of artificial intelligence (AI) quantitative parameters of pulmonary ground-glass nodules (GGN) in predicting the degree of infiltration. Methods A retrospective analysis of 168 consecutive patients with 178 GGNs in our hospital from October 2019 to May 2021 was performed, including 43 males and 125 females, aged 21-78 (55.76±10.88) years. Different lesions of the same patient were analyzed as independent samples. Totally, 178 GGNs were divided into two groups, a non-invasive group (24 adenocarcinoma in situ and 77 minimally invasive adenocarcinoma), and an invasive group (77 invasive adenocarcinoma). We compared the difference of AI quantitative parameters between the two groups, and evaluated predictive valve by receiver operating characteristic curve and binary logistic regression model. Results (1) Except for the gender (P=0.115), the other parameters, such as maximal diameter [15.10 (11.50, 21.60) mm vs. 8.90 (7.65, 11.15) mm], minimum diameter [10.80 (8.85, 15.20) mm vs. 7.40 (6.10, 8.95) mm], proportion of consolidation/tumor ratio [13.58% (1.61%, 63.76%) vs. 0.00% (0.00%, 0.67%)], mean CT value [–347.00 (–492.00, –101.50) Hu vs. –598.00 (–657.50, –510.00) Hu], CT maximum value [40.00 (–40.00, 94.50) Hu vs. –218.00 (–347.00, –66.50) Hu], CT minimum value [–584.00 (–690.50, –350.00) Hu vs. –753.00 (–786.00, –700.00) Hu], danger rating (proportion of high-risk nodules, 92.2% vs. 66.3%), malignant probability [91.66% (85.62%, 94.92%) vs. 81.81% (59.98%, 90.29%)] and age (59.93±8.53 years vs. 52.04±12.10 years) were statistically significant between the invasive group and the non-invasive group (all P<0.001). (2) The highest predictive value of a single quantitative parameter was the maximal diameter (area under the curve=0.843), the lowest one was the risk classification (area under the curve=0.627), the combination of two among the three parameters (maximal diameter, mean CT value, and consolidation/tumor ratio) improved the predictive value entirely. (3) Logistic regression analysis showed that maximal diameter and mean CT value both were the independent risk factor for predicting invasive adenocarcinoma. (4) When the threshold of v was 1.775%, the diagnostic sensitivity of invasive adenocarcinoma was 0.753 and the specificity was 0.851. Conclusion AI quantitative parameters can effectively predict the degree of infiltration of GGNs and provide a reliable reference basis for clinicians.
ObjectiveTo observe the diagnostic value of six classification intelligent auxiliary diagnosis lightweight model for common fundus diseases based on fundus color photography. MethodsA applied research. A dataset of 2 400 color fundus images from Nanjing Medical University Eye Hospital and Zhejiang Mathematical Medical Society Smart Eye Database was collected, which was desensitized and labeled by a fundus specialist. Of these, 400 each were for diabetic retinopathy, glaucoma, retinal vein occlusion, high myopia, age-related macular degeneration, and normal fundus. The parameters obtained from the classical classification models VGGNet16, ResNet50, DenseNet121 and lightweight classification models MobileNet3, ShuffleNet2, GhostNet trained on the ImageNet dataset were migrated to the six-classified common fundus disease intelligent aid diagnostic model using a migration learning approach during training as initialization parameters for training to obtain the latest model. 1 315 color fundus images of clinical patients were used as the test set. Evaluation metrics included sensitivity, specificity, accuracy, F1-Score and agreement of diagnostic tests (Kappa value); comparison of subject working characteristic curves as well as area under the curve values for different models. ResultCompared with the classical classification model, the storage size and number of parameters of the three lightweight classification models were significantly reduced, with ShuffleNetV2 having an average recognition time per sheet 438.08 ms faster than the classical classification model VGGNet16. All 3 lightweight classification models had Accuracy > 80.0%; Kappa values > 70.0% with significant agreement; sensitivity, specificity, and F1-Score for the diagnosis of normal fundus images were ≥ 98.0%; Macro-F1 was 78.2%, 79.4%, and 81.5%, respectively. ConclusionThe intelligent assisted diagnosis of common fundus diseases based on fundus color photography is a lightweight model with high recognition accuracy and speed; the storage size and number of parameters are significantly reduced compared with the classical classification model.
ObjectiveTo evaluate the diagnostic value of artificial intelligence (AI)-assisted diagnostic system for pulmonary cancer based on CT images.MethodsDatabases including PubMed, The Cochrane Library, EMbase, CNKI, WanFang Data and Chinese BioMedical Literature Database (CBM) were electronically searched to collect relevant studies on AI-assisted diagnostic system in the diagnosis of pulmonary cancer from 2010 to 2019. The eligible studies were selected according to inclusion and exclusion criteria, and the quality of included studies was assessed and the special information was identified. Then, meta-analysis was performed using RevMan 5.3, Stata 12.0 and SAS 9.4 softwares. The sensitivity, specificity, positive likelihood ratio, negative likelihood ratio and diagnostic odds ratio were pooled and the summary receiver operating characteristic (SROC) curve was drawn. Meta-regression analysis was used to explore the sources of heterogeneity.ResultsTotally 18 studies were included with 4 771 patients. Random effect model was used for the analysis due to the heterogeneity among studies. The results of meta-analysis showed that the pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, diagnosis odds ratio and area under the SROC curve were 0.87 [95%CI (0.84, 0.90)], 0.89 [95%CI (0.84, 0.92)], 7.70 [95%CI (5.32, 11.15)], 0.14 [95%CI (0.11, 0.19)], 53.54 [95%CI (30.68, 93.42)] and 0.94 [95%CI (0.91, 0.95)], respectively.ConclusionAI-assisted diagnostic system based on CT images has high diagnostic value for pulmonary cancer, and thus it is worthy of clinical application. However, due to the limited quality and quantity of included studies, above results should be validated by more studies.
In recent years, the computer science represented by artificial intelligence and high-throughput sequencing technology represented by omics play a significant role in the medical field. This paper reviews the research progress of the application of artificial intelligence combined with omics data analysis in the diagnosis and treatment of non-small cell lung cancer (NSCLC), aiming to provide ideas for the development of a more effective artificial intelligence algorithm, and improve the diagnosis rate and prognosis of patients with early NSCLC through a non-invasive way.
Artificial intelligence (AI) has been widely used in all walks of life, including healthcare, and has shown great application value in the auxiliary diagnosis of pulmonary nodules in the medical field. In the face of a large amount of lung imaging data, clinicians use AI tools to identify lesions more quickly and accurately, improving work efficiency, but there are still many problems in this field, such as the high false positive rate of recognition, and the difficulty in identifying special types of nodules. Researchers and clinicians are actively developing and using AI tools to promote their continuous evolution and make them better serve human health. This article reviews the clinical application and research progress of AI-assisted diagnosis of pulmonary nodules.
Evidence-based medicine is the methodology of modern clinical research and plays an important role in guiding clinical practice. It has become an integral part of medical education. In the digital age, evidence-based medicine has evolved to incorporate innovative research models that utilize multimodal clinical big data and artificial intelligence methods. These advancements aim to address the challenges posed by diverse research questions, data methods, and evidence sources. However, the current teaching content in medical schools often fails to keep pace with the rapidly evolving disciplines, impeding students' comprehensive understanding of the discipline's knowledge system, cutting-edge theories, and development directions. In this regard, this article takes the opportunity of graduate curriculum reform to incorporate real-world data research, artificial intelligence, and bioinformatics into the existing evidence-based medicine curriculum, and explores the reform of evidence-based medicine teaching in the information age. The aim is to enable students to truly understand the role and value of evidence-based medicine in the development of medicine, while possessing a solid theoretical foundation, a broad international perspective, and a keen research sense, in order to cultivate talents for the development of the evidence-based medicine discipline.
Compared with traditional medical devices, artificial intelligence medical devices face greater challenges in the process of clinical trials due to their related characteristics of artificial intelligence technology. This paper focused on the challenges and risks in each stage of clinical trials on artificial intelligence medical devices for assisted diagnosis, and put forward corresponding coping strategies, with the aim to provide references for the performance of high-quality clinical trials on artificial intelligence medical devices and shorten the research period in China.
ObjectiveTo explore the application of artificial intelligence in postoperative follow-up of day surgery patients, so as to establish an intelligent medical framework, promote the intelligent process of hospitals, and improve the management level of day surgery.MethodsThe artificial intelligence phonetic system was carried out by the Day Surgery Center, Renji Hospital, Shanghai Jiaotong University School of Medicine on June 1st, 2018. Through the system, the artificial intelligence voice system based on speech and semantic recognition technology was adopted to connect the data of the information center in the hospital to carry out postoperative follow-up of day surgery patients. We selected the 2 245 patients followed up by the artificial intelligence phonetic system from June 1st to November 30th 2018 (the AI follow-up group) and the 2 576 patients followed up by the traditional manual method from January 2nd to May 31st 2018 (the manual follow-up group), to compare the telephone connection rate, information collection rate, and call duration between them.ResultsThere was no statistically significant difference in telephone connection rate (85.70% vs. 86.68%) or information collection rate (98.86% vs. 98.48%) between the AI follow-up group and the manual follow-up group (P>0.05); but there was a statistically significant difference in call duration between the AI follow-up group and the manual follow-up group [(165.48±43.28) vs. (135.37±36.31) seconds, P<0.05], and the AI follow-up group had a longer call duration.ConclusionsThe application of artificial intelligence phonetic system in surgery has a good performance in call connection rate and information collection integrity. It plays an active role in improving efficiency, extending medical services and strengthening medical safety in the management of day surgery.
The development of the fifth generation mobile networks (5G) technology has brought great breakthroughs and challenges to clinical medicine and medical education. In the context of “5G + medicine”, the development of telemedicine, emergency rescue, intelligent analysis and diagnosis has opened up new horizons for clinical medicine. Facing the constant impact of high technology, the focus of medical education should be on the cultivation of students’ integrated medical view, critical thinking, communication abilities and skills, and creativity. The “5G + education” model will be presented by means of virtual reality, artificial intelligence, cloud computing and other technologies, providing a new direction for the development of medical education. This article summarizes the key points and prospects of medical education under 5G technology in order to provide a reference for the field of medical education to adapt to the changes in the 5G era.