Meta-Radiology 是由中南大学湘雅二医院主办，科爱公司发行的一本国际性同行评议（Peer-Review）和开放获取（Open Access）的医学影像领域英文学术期刊。本刊由中国科学院院士、中科院深圳先进技术研究院郑海荣教授担任名誉主编，中南大学副校长黎志宏教授担任主编，佐治亚大学杰出教授刘天明担任共同主编，中南大学湘雅二医院放射科主任刘军教授担任执行主编。Meta-Radiology 聚焦于医学影像交叉研究领域的最新研究成果，内容涵盖临床放射学、医学影像人工智能、分子影像、医学影像新技术、介入放射学、放射治疗等其他医学影像相关前沿科学研究，致力打造医学影像国际化学术交流平台，为推动及传播医学影像学领域最新研究进展提供一个崭新的国际化论坛。
![]()
Volume 2 Issue 2 2024
Research Article
Review Article
Editorial
Short Communication
Case Report
Guideline/Consensus
Meeting Reports
News & Announcements
Letter to the Editor
Perspective
https://www.keaipublishing.com/en/journals/meta-radiology/
https://www.editorialmanager.com/metrad/default2.aspx
www.sciencedirect.com/journal/meta-radiology
Meta-Radiology创刊初期2023年-2025年免收文章出版费（APC）
Meta-Radiology期刊第2卷第2期正式上线文章：
![]()
本期内容关键词：Diffusion probabilistic models, Score based generative modeling, MRI, Capsule network, Chest CT scan, Deep learning, Feature pyramid networks, Transfer learning, Attention mechanism, GPT–4, Artificial general intelligence, Knowledge diffusion, Interpret ability and explain ability, Societal influences, Governance, Deep learning, Soft tissue sarcomas, Transfer learning, Generative adversarial network, Algorithm, Artificial intelligence large models, Radiology, Progress, Challenges, Esophagus carcinoma, Lymph node metastasis, Texture analysis, Radiomics, Deep learning.
本期共发表7篇文章，包括1篇Editorial，2篇Article，4篇Review。本期作者分布：本期作者来自全球14所机构：伦敦国王学院圣托马斯医院、新加坡科技研究局、斯坦福大学、上海科技大学、湖南师范大学、中国科学院深圳先进技术研究院、中南大学湘雅二医院、中南大学深圳研究院、湖南大学、国防科技大学、杭州显盈科技团队、中国林业科学研究院、邛崃市医疗中心医院、重庆医科大学附属第二医院。![]()
Integrating multimodal and multiscale data: A new avenue in cancer research
With the help of massive advancements in artificial intelligence (AI) and the availability of an increasing amount of medical data, the past few decades have witnessed significant progression intelligent cancer research. Among those types of medical data fed into deep neural models, medical images generated by different radiology modalities are prominent. Driven by the urgent clinical needs, these medical images are trained to tackle specific tasks in the clinical pathway, such as scanning, reconstruction, diagnosis, prediction, follow-up, etc. These AI models have produced impressive achievements and efficiently help the clinicians to optimize the clinical decision making. However, most of the previous models are built on unimodal data and ignore the important complementary information between multimodal and multiscale data.This leads to substantial concerns about generalizability, robustness, and interpretability.
As we all know that the pathogenesis of cancer involves complex rearrangements at several levels, e.g., transcriptional, genetic, proteomic, and metabolic levels. Thus, in-depth analysis of the multimodal data is imperative, which allows the comprehensive characterization of cancer biological process and the dissection of tumor plasticity and heterogeneity. However, information derived from computed tomography (CT) or magnetic resonance (MR) images can only reveal the heterogeneity of tumors at the level of whole tumors or organs, and it is difficult to accurately evaluate the inherent biological characteristics such as tumor infiltrating lymphocytes (TILs), gene expression, and the arrangement of different cells. In recent years, multimodal and multiscale data integration becomes the prominencestrategy for investigators to explore the heterogeneity of tumors and pursue the superior model performance comprehensively and accurately.Taken together, by analysing medical multimodal and multiscale data, investigators can innovate the avenue of cancer research in terms ofidentifying genetic susceptibilities, drug response patterns, predictive biomarkers, etc., thus laying the groundwork for precision medicine. The continual innovation of AI techniques will likely accelerate the process of this avenue and potentially improve the efficiency of prevention and treatment of cancer. We invite Articles that aim to spotlight the multimodal and multiscale medical data in cancer research area with the potential to aid the entire smart healthcare. We hope that Meta-Radiology will serve as a platform for innovations, addressing potential limitations and challenges associated with multimodal and multiscale data analysis. 得益于人工智能 (artificial intelligence, AI) 的飞速发展和大量医疗数据的出现，过去几十年来，癌症智能研究取得了重大进展。在输入到深度神经模型的医疗数据类型中，由不同放射学模式生成的医学图像尤为突出。在临床迫切需求的推动下，这些医学图像经过训练，可以解决临床路径中的特定任务，例如扫描、重建、诊断、预测、随访等。这些 AI 模型取得了令人瞩目的成就，并有效地帮助临床医生优化临床决策。然而，大多数先前的模型都是基于单模态数据构建的，忽略了多模态和多尺度数据之间的重要互补信息。这引起了人们对通用性、稳健性和可解释性的极大担忧。众所周知，癌症的发病机制涉及多个水平的复杂重排，例如转录、遗传、蛋白质组和代谢水平。因此，对多模态数据的深入分析至关重要，这有助于全面表征癌症的生物学过程并解剖肿瘤的可塑性和异质性。然而，来自计算机断层扫描 (computed tomography, CT) 或磁共振 (magnetic resonance, MR) 图像的信息只能揭示肿瘤在整个肿瘤或器官水平上的异质性，很难准确评估肿瘤浸润淋巴细胞(tumor infiltrating lymphocytes, TIL)、基因表达和不同细胞的排列等固有生物学特性。近年来，多模态和多尺度数据整合成为研究者全面准确地探索肿瘤异质性和追求卓越模型性能的主要策略。总之，通过分析医学多模态和多尺度数据，研究人员可以在识别遗传易感性、药物反应模式、预测性生物标志物等方面创新癌症研究途径，从而为精准医疗奠定基础。人工智能技术的不断创新可能会加速这一途径的进程，并有可能提高癌症预防和治疗的效率。我们诚邀关注癌症研究领域的多模态和多尺度医学数据的文章，这些文章有可能为整个智能医疗提供帮助。我们希望Meta-Radiology能够成为一个创新平台，解决与多模态和多尺度数据分析相关的潜在限制和挑战。Zheng H. Integratingmultimodal and multiscale data: a new avenue in cancer research[J]. Meta-Radiology, 2024: 100083.
扫描获取全文（网页全文二维码）
![]()
MCFCN: Multi-scale capsule-weighted fusion classification network for lung disease classification based on chest CT scans
Aim and scope: This paper aims to propose a Multi-scale Capsule-weighted Fusion Classification Network (MCFCN), a classification model for automatic diagnosis of lung lesions by CT scanning.Background: The automatic diagnosis of lung lesions based on chest CT scans plays a crucial role in assisting doctors to identify suspicious cases quickly and accurately. However, existing methods struggle to differentiate lesions with similar morphologies, and current feature extraction techniques lack the ability to effectively highlight small-scale targets in a large-scale environment, leading to incomplete extraction of subtle features and ultimately compromising the classification performance.Method: The MCFCN employs a dynamic routing clustering algorithm to emphasize small-scale features, preventing feature loss. Additionally, a scale difference fusion network is utilized to extract precise position scaling parameters by incorporating weighted fusion of information from different scales.Results: MCFCN achieves an accuracy of 99.41% for COVID-19 classification, 93.33% for CAP classification, and 100% for Normal classification, with an overall accuracy of 98.36%.Conclusion: Experimental results on the target dataset demonstrate that MCFCN outperforms state-of-the-art methods. In the future, this model can be further explored and optimized to enhance its application value in clinical practice.
![]()
MCFCN模型
基于肺部CT的肺部病变自动诊断技术，可以帮助医生快速、准确地发现可疑病例，为疾病的及时救治和预防发挥重要作用。但现有方法无法区分具有近似形态的病变，同时现有的特征提取方法缺乏在大尺度环境中凸显小尺度目标的手段，无法实现微小特征高效提取，这会导致病变特征提取不完整，最终削弱分类效果。本文主要介绍一种用于自动检测肺部CT图像的多尺度胶囊加权融合分类网络(Multi-scale Capsule-weighted Fusion Classification Network, MCFCN)。该网络由多尺度动态路由特征提取网络、多尺度切片融合、三维加权融合等部分组成。针对大尺度医学图像中微小病变区域关注度不高的问题，我搭建了多尺度特征加权融合分类网络，用动态路由结构对不同尺度的特征图进行增强和削弱，使用尺度差异融合网络在融合过程中得到高效的位置放缩参数；并且，针对现有数据集中数据量太少从而影响了模型的学习效果，无法区分具有近似形态病变的问题，本文提出一种迁移学习训练方法，使用高区分度数据集对模型特征提取结构进行针对性训练，使得模型对多个类别的病变都有强大的分辨能力，增强了网络的分类能力。我们评估了MCFCN在COVID-CT-MD数据集上的性能，并与几种其他主流算法进行比较，MCFCN模型优于现有的SOTA方法，特别是灵敏度达到了99.63%，这意味着模型进一步降低了误检的可能性。此外，特异性为98.73%，模型也降低了漏检的可能性。Liu A, Liu S, Wen C. MCFCN: Multi-scale capsule-weighted fusion classification network for lung disease classification based on chest CT scans[J]. Meta-Radiology, 2024, 2(2): 100070.
![]()
扫描获取全文（网页全文二维码）
#文章详细内容链接：湖南师范大学温翠红团队深入研究提出一种基于MCFCN的肺部CT图像自动诊断算法
![]()
The general intelligence of GPT-4, its knowledge diffusive and societal influences, and its governance
Recent breakthroughs in artificial intelligence (AI) research include advancements in natural language processing (NLP) achieved by large language models (LLMs), and; in particular, generative pre-trained transformer (GPT) architectures. The latest GPT developed by OpenAI, GPT-4, has shown remarkable intelligence across various domains and tasks. It exhibits capabilities in abstraction, comprehension, vision, computer coding, mathematics, and more, suggesting it to be a significant step towards artificial general intelligence (AGI), a level of AI that possesses capabilities similar to human intelligence. This paper explores this AGI, its knowledge diffusive and societal influences, and its governance. In addition to coverage of the major associated topics studied in the literature, and making up for their loopholes, we scrutinize how GPT-4 can facilitate the diffusion of knowledge across different areas of science by promoting their interpretability and explainability (IE) to inexperts. Where applicable, the topics are also accompanied by their specificpotential implications on medical imaging. 人工智能 (artificial intelligence, AI) 研究的最新突破包括通过大型语言模型 (large language models, LLM) 实现的自然语言处理 (natural language processing, NLP) 方面的进步，特别是生成式预训练转换器 (generative pre-trained transformer, GPT) 架构。基于OpenAI 开发的最新GPT，GPT-4 在各个领域和任务中表现出了卓越的智能。它展现了抽象、理解、视觉、计算机编码、数学等方面的能力，表明它是迈向通用人工智能 (artificial general intelligence, AGI) 的重要一步，通用人工智能是一种具有与人类智慧相似的 AI 级别。本文探讨了通用人工智能的知识传播和社会影响及其管理制度。除了涵盖文献中研究的主要相关主题并弥补其漏洞外，我们还仔细研究了GPT-4如何通过提高其对非相关领域专家的可解释性来促进不同科学领域的知识传播。在适用的情况下，这些主题还伴随着它们对医学成像的潜在影响。Yekta M M J. The general intelligence of GPT–4, its knowledge diffusive and societal influences, and its governance[J]. Meta-Radiology, 2024, 2(2): 100078.
![]()
扫描获取全文（网页全文二维码）
![]()
A survey of emerging applications of diffusion probabilistic models in MRI
作者：Yuheng Fan, Hanxi Liao, Shiqi Huang, Yimin Luo, Huazhu Fu, Haikun Qi*.
内容简介
Diffusion probabilistic models (DPMs) which employ explicit likelihood characterization and a gradual sampling process to synthesize data, have gained increasing research interest. Despite their huge computational burdens due to the large number of steps involved during sampling, DPMs are widely appreciated in various medical imaging tasks for their high-quality and diversity of generation. Magnetic resonance imaging (MRI) is an important medical imaging modality with excellent soft tissue contrast and superb spatial resolution, which possesses unique opportunities for DPMs. Although there is a recent surge of studies exploring DPMs in MRI, a survey paper of DPMs specifically designed for MRI applications is still lacking. This review article aims to help researchers in the MRI community to grasp the advances of DPMs in different applications. We first introduce the theory of two dominant kinds of DPMs, categorized according to whether the diffusion time step is discrete or continuous, and then provide a comprehensive review of emerging DPMs in MRI, including reconstruction, image generation, image translation, segmentation, anomaly detection, and further research topics. Finally, we discuss the general limitations as well as limitations specific to the MRI tasks of DPMs and point out potential areas that are worth further exploration.![]()
扩散概率模型中的扩散过程和逆过程
扩散概率模型 (Diffusion probabilistic models, DPM) 采用显式似然表征和逐步采样过程来合成数据，引起了越来越多学者的研究兴趣。尽管由于采样过程中涉及大量步骤，DPM的计算负担巨大，但DPM因其高质量和生成多样性而在各种医学成像任务中受到广泛赞赏。磁共振成像 (magnetic resonance imaging, MRI) 是一种重要的医学成像方式，具有出色的软组织对比度和空间分辨率，为 DPM 提供了独特的机会。尽管目前有大量研究探索MRI中的DPM，但为MRI应用设计的DPM的调查论文仍然缺乏。这篇综述旨在帮助MRI相关专业的研究人员了解及掌握DPM在不同应用中的进展。我们首先介绍了两种主要DPM的理论，根据扩散时间步长是离散的还是连续的进行分类，然后全面回顾了MRI中新兴的DPM，包括重建、图像生成、图像转换、分割、异常检测以及进一步的研究主题。最后，我们讨论了DPM的一般局限性以及特定于MRI任务的局限性，并指出了值得进一步探索的潜在领域。Fan Y, Liao H, Huang S, et al. A survey of emerging applications of diffusion probabilistic models in mri[J]. Meta-Radiology, 2024: 100082.
![]()
扫描获取全文（网页全文二维码）
![]()
Deep learning-based artificial intelligence for assisting diagnosis, assessment and treatment in soft tissue sarcomas
作者：Ruiling Xu, Jinxin Tang, Chenbei Li, Hua Wang, Lan Li, Yu He, Chao Tu*, Zhihong Li*.
内容简介
Radiation therapy (RT) is currently the main clinical treatment of tumors. Soft tissue sarcomas (STSs) represent a group of heterogeneous mesenchymal tumors of which are generally classified as per the histopathology. Despite being rare in incidence and prevalence, STSs are usually correlated with unfavorable prognosis and high mortality rate. Early and accurate diagnosis of STSs are critical in clinical management of STSs. Deep learning (DL) refers to a subtype of artificial intelligence that has been adopted to assist healthcare professionals to optimize personalized treatment for a given situation, particularly in image analysis. Recently, emerging studies have demonstrated that application of DL based on medical images could substantially improve the accuracy and efficiency of clinicians to the identification, diagnosis, treatment, and prognosis prediction of STSs, and thereby facilitating the clinical decision-making. Herein, we aimed to extensively summarize the recent applications of DL-based artificial intelligence in STSs from the aspects of data acquisition, algorithm, and model establishment. Besides, the reinforcement of the model by transfer learning and generative adversarial network (GAN) for data augmentation has also been elaborated. It is worth noting that high-quality data with accurate annotations, as well as optimized algorithmic performance are pivotal in the clinical application of DL in STSs. ![]()
软组织肉瘤中基于深度学习的AI工作原理
软组织肉瘤(soft tissue sarcomas, STSs)是一组异质性高的间充质恶性肿瘤，一般按组织病理学进行分类。尽管临床上STSs的发病率和患病率很低，但通常与不良预后和高死亡率相关。STSs的早期、准确诊断是临床治疗的关键。深度学习(Deep learning，DL)是人工智能的一个子类型，已应用于帮助医疗专业人员针对特定情况优化个性化治疗。首先，准确的诊断和分类是医学过程中最关键的一步，DL模型表现出优异的诊断能力和更加精细的分类能力，对于指导临床医生的后续治疗非常有帮助；其次，精准治疗是医学过程中核心步骤，DL模型可以帮助指导敏感性药物的使用，避免耐药性的出现；最后，临床结果的预测是临床医生和患者都渴望知道的，因为它可以用来指导早期临床干预。在这方面，DL模型可以通过分析临床数据和影像、病理图像数据来获得准确的结果并进行预测。因此，将深度学习与影像学、病理学和临床标准参数结合使用有望提高疾病诊断的准确性和效率。近年来，一系列研究表明，基于医学图像的深度学习的应用可以显著增加临床医生对STSs的识别、诊断、治疗和预后预测的准确性，且效率明显提升，从而促进临床决策。本文从数据获取、算法和模型建立三个主要方面详细总结了基于深度学习的人工智能在STSs中的最新应用进展。此外，还详细阐述了通过迁移学习和生成对抗网络(Generative adversarial network, GAN)进行数据增强的模型强化。总的来说，具有准确注释的高质量的数据、以及性能优化的算法对于深度学习在STSs中的临床应用至关重要。Xu R, Tang J, Li C, et al. Deep Learning-based Artificial Intelligence for Assisting Diagnosis, Assessment and Treatment in Soft Tissue Sarcomas[J]. Meta-Radiology, 2024: 100069.扫描获取全文（网页全文二维码）
![]()
Opportunities and challenges in the application of large artificial intelligence models in radiology
作者：Liangrui Pan, Zhenyu Zhao, Ying Lu, Kewei Tang, Liyong Fu*, Qingchun Liang*, Shaoliang Peng*.
内容简介
Influenced by ChatGPT, artificial intelligence (AI) large models have witnessed a global upsurge in large model research and development. As people enjoy the convenience by this AI large model, more and more large models in subdivided fields are gradually being proposed, especially large models in radiology imaging field. This article first introduces the development history of large models, technical details, workflow, working principles of multimodal large models and working principles of video generation large models. Secondly, we summarize the latest research progress of AI large models in radiology education, radiology report generation, applications of unimodal and multimodal radiology. Finally, this paper also summarizes some of the challenges of large AI models in radiology, with the aim of better promoting the rapid revolution in the field of radiography.![]()
受ChatGPT影响，人工智能 (artificial intelligence, AI) 大模型掀起了全球大模型研发热潮。随着人们享受人工智能大模型带来的便利，越来越多的细分领域的大模型逐渐被提出，特别是放射成像领域的大模型。本文首先介绍了大模型的发展历史、技术细节、工作流程、多模态大模型的工作原理以及视频生成大模型的工作原理。其次，总结了人工智能大模型在放射学教育、放射学报告生成、单模态和多模态放射学应用等方面的最新研究进展。最后，本文还总结了大型AI模型在放射学领域面临的一些挑战，旨在更好地推动放射学领域的快速革命。
Pan L, Zhao Z, Lu Y, et al. Opportunities and challenges in the application of large artificial intelligence models in radiology[J]. Meta-Radiology, 2024: 100080.扫描获取全文（网页全文二维码）
#本期封面文章
#文章详细内容链接：湖南大学信息科学与工程学院彭绍亮教授团队综述了人工智能大模型在放射学领域的最新研究进展
![]()
Review and prospects of new progress in intelligent imaging research on lymph node metastasis in esophageal carcinoma
Esophagus carcinoma (EC) ranks sixth in cancer-related mortality and seventh in terms of morbidity worldwide, and radical esophagectomy is considered as the basis of comprehensive treatment for locally advanced EC. Accurate preoperative determination of lymph node status is critical for treatment decision-making, assessment of survival time and life quality of patients after surgery. However, the rate of misdiagnosis and missed diagnosis of metastatic lymph nodes by traditional imaging methods is high. With the development of artificial intelligence technology and medical image digitization, medical image analysis methods based on artificial intelligence have brought new ideas to the diagnosis and research of lymph node metastasis secondary to EC. At present, texture analysis, radiomics and deep learning are the most widely used methods. These technologies extract and analyze quantitative features from traditional medical images to provide biological information such as tumor characteristics and heterogeneity to guide clinical practice. Therefore, this review mainly introduces and discusses the current status of imaging research on lymph node metastasis in patients with EC based on texture analysis, radiomics and deep learning, and prospects the important research directions in the future with a view to improving the diagnostic capability of lymph node metastasis in patients with EC in China.![]()
影像组学和深度学习操作流程对比示意图
目前传统的影像学检查手段对转移性淋巴结的漏诊率及误诊率较高。随着人工智能技术和数字化影像的发展，基于人工智能 (Artificial intelligence, AI) 的医学影像图像分析方法为食管癌淋巴结转移 (Lymph node metastasis, LNM) 的诊断研究带来了新思路。目前研究应用最多的主要包括纹理分析、影像组学和深度学习，这些技术从医学影像中提取更加细微的特征，更深入地分析食管癌的异质性，进而指导临床实践。但目前基于人工智能的影像研究还存在很多局限性。随着AI的快速发展，各种AI相关技术在医学领域发挥了越来越重要的作用。而在医学影像中研究应用较多的是机器学习 (Machine learning, ML)。机器学习包括最早的传统AI算法，如基于支持向量机、随机森林、朴素贝叶斯等的影像组学，其中纹理分析是影像组学技术之一。这些技术之后又出现了新的复杂的深度学习算法 (Deep learning, DL)。近年来，影像图像纹理分析已应用于肿瘤的诊断、治疗及预后评估等多个领域。影像纹理分析对食管癌LNM具有较好的诊断价值，但目前缺乏公认的最优纹理参数及标准值作为诊断标准。影像组学可在术前无创地预测食管癌LNM，从而克服目前传统影像学以淋巴结大小判断有无LNM的局限性，但尚缺乏统一的操作流程来确保其分析结果的准确性。而目前关于DL相关的医学研究相对较少，目前还属于新兴领域，但基于其强大的数据分析能力，DL在医学研究中具有广阔应用前景。Gao D, Wu Y, Chen T. Review and prospects of new progress in intelligent imaging research on lymph node metastasis in esophageal carcinoma[J]. Meta-Radiology, 2024: 100081.扫描获取全文（网页全文二维码）
#文章详细内容链接：重庆医科大学附属第二医院放射科陈天武团队综合述评并展望了食管癌淋巴结转移的智能影像研究进展
