2025年IMIC实验室学术年会暨MICS东部地区学术沙龙讲者介绍

付华柱 研究员

新加坡科技研究局 / A*STAR

报告题目/Title:《面向医学视觉-语言模型中可靠性和可解释性问题的研究》/ Addressing Reliability and Explainability in Medical Vision-Language Models

报告摘要：医学视觉-语言模型（VLMs）在临床应用领域已展现出巨大潜力，但在实际医疗场景中的部署仍受两大关键挑战制约：可靠性问题与可解释性不足。本报告将介绍我们近期的研究成果，通过不确定性感知建模与基于事实的解释生成两种技术路径，同步应对上述两大挑战。这些进展推动了可信医疗人工智能系统的发展，使其能够提供可靠的诊断支持，并具备临床接受度和患者安全所需的透明推理过程。该研究的核心贡献包括：首先，构建了基于不确定性的可靠性机制，用于实现分布外检测与幻觉预防；其次，设计了基于事实的解释框架，用于提升临床可解释性；然后，基于多个医学影像中心完成了实验验证；最后，通过原则性的推理方法，在降低训练需求的同时实现性能提升。

Abstract: Medical Vision-Language Models (VLMs) have demonstrated substantial promise in clinical applications, yet their deployment in real-world healthcare settings remains constrained by two critical challenges: reliability concerns and limited explainability. This talk presents our recent works addressing both dimensions through uncertainty-aware modeling and grounded explanation generation. These developments advance trustworthy medical AI systems that provide reliable diagnostic support with transparent reasoning processes essential for clinical acceptance and patient safety. Key contributions include uncertainty-based reliability mechanisms for out-of-distribution detection and hallucination prevention, grounded explanation frameworks for enhanced clinical interpretability, empirical validation across multiple medical imaging domains, and performance improvements with reduced training requirements through principled reasoning approaches.

个人简介：付华柱博士，现任新加坡科技研究局 (A*STAR) 高性能计算研究所 (IHPC) 主任研究员。主要研究方向为AI for Healthcare 以及 Trustworthy AI 等。至今已在 Nature 子刊，Cell 子刊，IEEE TPAMI 等期刊和会议上发表论文 200 余篇，Google Scholar 引用 3 万余次。入选科睿唯安（Clarivate）全球“高被引科学家”。现担任 IEEE TMI，IEEE TNNLS，和 IEEE JBHI 等期刊编委。

Brief Bio: Dr. Huazhu Fu is a Senior Research Fellow at the Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A*STAR), Singapore. His primary research focuses on AI for Healthcare and Trustworthy AI. To date, he has published over 200 papers in journals and conferences such as Nature sub-journals, Cell sub-journals, and IEEE TPAMI, with over 30,000 citations on Google Scholar. He has been recognized as a "Highly Cited Researcher" by Clarivate Analytics. Dr. Zhu currently serves as an editorial board member for several prestigious journals, including IEEE TMI, IEEE TNNLS, and IEEE JBHI.

---

陈阳 教授

东南大学 / Southeast University

报告题目/Title:《基于多模态大模型的脑卒中智能化诊疗》/ Large Models for Intelligent Stroke Diagnosis and Treatment

报告摘要：脑卒中作为全球致残及致死率最高的疾病之一，其早期精准诊断与个体化治疗对于显著改善患者预后具有决定性意义。在临床实践中，梗死核心区与缺血半暗带的准确识别，是制定最优治疗策略、把握救治窗口期的两大关键环节。梗死核心区代表不可逆的脑组织损伤，而缺血半暗带则为具有挽救潜力的区域，对二者的精确界定直接决定溶栓及取栓等再灌注治疗的适应性与时效性。然而，传统依赖医师主观经验的判读方法在梗死核心区与半暗带的识别上存在一定局限，难以充分挖掘多模态影像（如CT、MRI及灌注成像等）中蕴含的丰富且互补的病理信息。多模态大模型驱动的人工智能技术，能够深度融合多源异质数据，自动提取并综合分析关键影像特征，实现卒中病灶的高精度、自动稳定的定量识别，有望推动脑卒中诊疗的智能化和精准化进程；同时，该技术在低场磁共振成像领域也展现出良好的鲁棒性和稳定性，有望进一步推广脑卒中在床旁车载等移动便携场景下的早期风险预警与评估。

Abstract: Stroke is one of the diseases with the highest disability and mortality rates worldwide. Its early accurate diagnosis and individualized treatment are crucial for significantly improving patient prognosis. In clinical practice, the accurate identification of the infarct core and ischemic penumbra are two key links in formulating optimal treatment strategies and seizing the treatment window. The infarct core represents irreversible brain tissue damage, while the ischemic penumbra is a region with salvage potential. The precise definition of both directly determines the suitability and timeliness of reperfusion therapies such as thrombolysis and thrombectomy. However, traditional interpretation methods that rely on physicians' subjective experience have certain limitations in identifying the infarct core and penumbra, making it difficult to fully tap into the rich and complementary pathological information contained in multimodal imaging (such as CT, MRI, and perfusion imaging). Artificial intelligence technology driven by multimodal large models can deeply integrate multi-source heterogeneous data, automatically extract and comprehensively analyze key imaging features, and achieve high-precision, automatic, and stable quantitative identification of stroke lesions. This is expected to promote the intelligent and precise development of stroke diagnosis and treatment. Meanwhile, this technology also shows good robustness and stability in the field of low-field MRI, and is expected to further promote the early risk warning and assessment of stroke in mobile and portable scenarios such as bedside and vehicle-mounted settings.

个人简介：陈阳，东南大学教授，国家杰出青年科学基金获得者。深耕医学人工智能与医学图像处理前沿技术，致力于将计算机科学技术应用于视觉重建、医学图像成像和计算机辅助诊断分析，论文发表在包括IEEE Transactions on Medical Imaging、Medical Image Analysis、IEEE Journal of Biomedical and Health Informatics等人工智能与医学影像处理领域顶刊。授权国家发明专利 11 项，获得吴文俊人工智能科技进步奖二等奖，中国体视学学会科技进步二等奖，中国图象图形学会科学技术奖一等奖和山东省科技进步一等奖。此外，申请人还承担国家杰出青年科学基金项目、国家重大科学研究计划、863重点项目等多项国家级课题，担任中国生物医学工程学会医学图像信息与控制分会副主任委员等。

Brief Bio: Dr. Yang Chen is a Professor at Southeast University and a recipient of the National Science Fund for Distinguished Young Scholars. He has long been engaged in cutting-edge technologies in medical artificial intelligence (AI) and medical image processing, and is committed to applying computer science and technology to visual reconstruction, medical image imaging, and computer-aided diagnostic analysis. His papers have been published in top journals in the fields of AI and medical image processing, including IEEE Transactions on Medical Imaging, Medical Image Analysis, and IEEE Journal of Biomedical and Health Informatics. He holds 11 authorized national invention patents and has received multiple awards, such as the Second Prize of Wu Wenjun Artificial Intelligence Science and Technology Progress Award, the Second Prize of Science and Technology Progress Award of Chinese Society for Stereology, the First Prize of Science and Technology Award of Chinese Society of Image and Graphics, and the First Prize of Shandong Provincial Science and Technology Progress Award. In addition, Dr. Chen undertakes a number of national-level projects, including projects supported by the National Science Fund for Distinguished Young Scholars, the National Major Scientific Research Program, and the Key Projects of the 863 Program. He also serves as the Deputy Director of the Medical Image Information and Control Branch of the Chinese Society of Biomedical Engineering.

---

王连生 教授

厦门大学 / Xiamen University

报告题目/Title: 高效计算病理/ Efficient Computational Pathology

报告摘要：随着大模型、生成式等人工智能算法的快速发展和代码开源，以及公开和私有病理切片数据的急剧增加，计算病理领域的研究突飞猛进。本次报告将主要介绍课题组在计算病理方面的最新研究进展和应用，主要包括疾病诊断、预后生存、多模态等方面的内容。

Abstract: With the rapid development of large models and generative AI algorithms, together with the increasing availability of open-source resources and large-scale pathology slide datasets, computational pathology has entered a period of rapid advancement. Achieving efficient and accurate pathology analysis under such data- and model-intensive conditions has become a key challenge in intelligent healthcare. This talk will present the group’s recent progress and applications in computational pathology, covering disease diagnosis, prognosis prediction, and multimodal pathology analysis. By exploring model efficiency, data utilization, and cross-modal fusion, this research aims to advance computational pathology toward greater efficiency, intelligence, and clinical applicability.

个人简介：王连生，现为厦门大学信息学院教授，医学院双聘教授，博士生导师，数字福建健康医疗大数据研究所副所长，福建省医学会放射学分会AI学组副组长，海医会智能医学影像与信息化专委会副主任，厦门大学医学院医学人工智能研究院负责人，MICS主席。长期从事医学影像处理研究，主持和参与多项科研项目，包括国家自然科学基金仪器专项、科技部科技创新2030重大项目、国家重点研发项目、国家自然科学基金面上和青年项目等，发表包括Nature Machine Intelligence、Nature Communications、Cell Reports Methods、Cell Patterns、人工智能顶会CVPR/AAAI等相关研究论文100余篇，获得腾讯犀牛鸟科研奖、CSPE Young Investigator、福建省科技进步二等奖、2023年厦门大学田昭武交叉学科一等奖，带领团队先后11次在国际医学影像比赛中获得冠军。

Brief Bio: Liansheng Wang is a Professor at the School of Informatics, jointly appointed by the School of Medicine, Xiamen University, and a doctoral supervisor. He currently serves as Deputy Director of the Digital Fujian Institute of Big Data for Health and Medicine, Director of the Institute of Medical Artificial Intelligence at Xiamen University, and Chair of MICS. He is also the Vice Chair of the AI Group of the Radiology Branch of the Fujian Medical Association and the Vice Chair of the Intelligent Medical Imaging and Informatics Committee of the Marine Medical Association. His research focuses on medical image analysis, where he has led and participated in multiple national and provincial projects, including the NSFC Instrumentation Program, the National Key R&D Program, and the Ministry of Science and Technology's "AI 2030" Major Project. He has published over 100 papers in top-tier journals and conferences, such as Nature Machine Intelligence, Nature Communications, Cell Reports Methods, Cell Patterns, CVPR, and AAAI. His honors include the Tencent Rhino-Bird Research Award, CSPE Young Investigator Award, the Fujian Provincial Science and Technology Progress Award (Second Prize), and the 2023 Xiamen University Tian Chaowu Interdisciplinary Award. His team has achieved 11 championships in international medical imaging competitions.

---

杨冠羽 教授

东南大学 / Southeast University

报告题目/Title: 基于深度学习的心血管影像智能分析算法研究/ Research on intelligent analysis algorithms for cardiovascular imaging based on deep learning

报告摘要：近年来，深度学习技术在医学图像分析领域进展迅速。但是，由于心血管影像模态多、维度高、标注难等特点，使得依靠大量精细标注图像监督学习的范式在心血管图像计算中面临巨大挑战。为此，近年我们结合临床实际需求，提出了多种融合先验知识的深度学习血管影像AI模型，在提高任务精度的同时降低了模型训练对精细标注数据的依赖。

Abstract: In recent years, deep learning technology has advanced rapidly in the field of medical image analysis. However, due to the characteristics of cardiovascular imaging, such as multiple modalities, high dimensionality, and difficulty in annotation, the paradigm of supervised learning relying on large amounts of meticulously annotated images faces significant challenges in cardiovascular image computing. To address this, in recent years, we have proposed various deep learning-based vascular imaging AI models that integrate prior knowledge, tailored to clinical needs. These models not only improve task accuracy but also reduce the dependency on finely annotated data for model training.

个人简介：杨冠羽，东南大学计算机科学与工程学院、软件学院、人工智能学院副院长、教授、博士生导师，IEEE高级会员。江苏省计算机学会理事，江苏省研究型医学学会结构性心脏病专业委员会常委。江苏省医学信息处理国际合作联合实验室副主任。东南大学生物医学工程专业博士、法国雷恩一大信号与图像处理专业博士、荷兰莱顿大学医学中心（LUMC，Leiden University）博士后。长期从事医学人工智能、图像处理与分析、计算机辅助诊断与手术方面的研究。承担国自然“重大疾病智慧诊疗”专项、国自然面上、国家重点研发计划课题、国家科技重大专项课题等十余项。发表包括IEEE TPAMI，IEEE TIP、IEEE TMI、Med Image Anal、CVPR、ICCV、IJCAI、MICCAI等期刊和会议在内的论文80余篇，授权国家发明专利12项。曾获得教育部自然科学二等奖、江苏省医学科技奖二等奖等。

Brief Bio: Guanyu Yang is a Professor and Doctoral Supervisor, currently serving as the Associate Dean of the School of Computer Science and Engineering / School of Software / School of Artificial Intelligence at Southeast University. He is a Senior Member of IEEE. He also holds the position of Deputy Director of the Jiangsu International Joint Laboratory for Medical Information Processing. Dr. Yang earned his Ph.D. in Biomedical Engineering from Southeast University and a Ph.D. in Signal and Image Processing from Université de Rennes I (France). He subsequently conducted postdoctoral research at the Leiden University Medical Center (LUMC) in the Netherlands. His long-term research focuses on medical artificial intelligence, image processing and analysis, and computer-aided diagnosis and surgery. He has undertaken more than ten national-level research projects, including the NSFC "Intelligent Diagnosis and Treatment for Major Diseases" Special Program, NSFC General Programs, key projects under the National Key R&D Program, and projects under the National Science and Technology Major Project. He has published over 80 papers in prestigious journals and conferences, including IEEE TPAMI, IEEE TIP, IEEE TMI, Medical Image Analysis, CVPR, ICCV, IJCAI, and MICCAI, and has been granted 12 national invention patents.

---

王猛 研究员

新加坡国立大学 / National University of Singapore

报告题目/Title: 面向眼科的可信且临床友好型人工智能研究/ Towards Trustworthy and Clinician-Friendly AI for Ophthalmology

报告摘要：人工智能（AI）正在迅速推动眼科诊断的发展，但其在临床中的广泛应用仍面临可靠性与可用性方面的挑战。本报告将介绍我们近期在构建可信且实用的眼科人工智能系统方面的一系列研究工作。这些研究包括基于不确定性感知与开放集识别的模型，用于实现更安全的临床决策支持；以及一种无需训练、面向临床医生的人工智能平台，旨在促进AI技术在真实医疗环境中的无缝落地。上述创新共同推动了AI从研究走向临床实践的转化，促进其在日常眼科诊疗中的可靠应用。

Abstract: Artificial intelligence (AI) is rapidly advancing ophthalmic diagnostics, yet its clinical adoption still faces challenges of reliability and usability. This talk will present a series of our recent works aimed at developing trustworthy and practical AI systems for ophthalmology. These include uncertainty-aware and open-set models that enable safer clinical decision-making, as well as a training-free, clinician-friendly AI platform designed for seamless integration into real-world practice. Together, these innovations strive to bridge the gap between research and clinical application, advancing the reliable use of AI in everyday ophthalmic care.

个人简介：王猛博士，现任新加坡国立大学高级研究员（Senior Research Fellow）。主要从事人工智能与医学影像分析方法研究，研究方向涵盖计算机视觉、医学图像处理、医学影像大模型及可信人工智能等领域。现担任《IEEE Journal of Biomedical and Health Informatics》及《Frontiers in Medicine》等期刊的客座编辑。迄今已发表学术论文60余篇，包括 Nature Communications、Cell Reports Medicine、npj Digital Medicine 和 IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 等国际顶级期刊，以及 CVPR 和 MICCAI 等国际顶级会议，并参与撰写专著 Federated Learning for Medical Imaging。

Brief Bio: Dr. Meng Wang is a Senior Research Fellow at the National University of Singapore. His research focuses on artificial intelligence and medical image analysis, encompassing computer vision, medical image processing, foundation models for medical imaging, and trustworthy AI. He serves as a guest editor for journals such as IEEE Journal of Biomedical and Health Informatics and Frontiers in Medicine. To date, he has published over 60 papers in leading journals and conferences, including Nature Communications, Cell Reports Medicine, npj Digital Medicine, IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), CVPR, and MICCAI, and has co-authored the book Federated Learning for Medical Imaging.

---

周毅 副教授

东南大学 / Southeast University

报告题目/Title: 基于可持续可推理基础模型的眼底图像辅诊/ Fundus Image Diagnosis: Towards An Incremental and Reasoning Foundation Model

报告摘要：眼底疾病已成为全球不可逆性致盲的重要原因，基于AI技术的眼底图像辅助诊断具有重要的研究意义。本报告围绕AI模型在眼底图像辅诊领域中的发展，从模型的通用泛化性、可持续性、可推理性等问题出发，介绍从专用模型时代（特定任务专用模型）、到前基础模型时代（预训练基础模型）、再到后基础模型时代（多模态推理基础模型）研究范式迁变中眼底图像辅诊模型面临的挑战和关键方法技术，其中的诸多思考与代表性工作也同样映射出其他类似医学图像辅诊场景的研究发展变化历程。

Abstract: Abstract:Fundus diseases have become a major cause of irreversible blindness worldwide, making AI-based fundus image-assisted diagnosis of significant research importance. This talk focuses on the development of AI models in the field of fundus image-assisted diagnosis, addressing issues such as model generalization, sustainability, and inference capabilities. It introduces the challenges and key methodologies faced by fundus image-assisted diagnostic models during the paradigm shift from the era of dedicated models (task-specific models), to the pre-foundation model era (pre-trained foundation models), and then to the post-foundation model era (multimodal reasoning models). Many of the reflections and representative works presented also reflect the research and development evolution of other similar medical image-assisted diagnostic scenarios.

个人简介：周毅，东南大学计算机科学与工程学院副教授，博士生导师。研究领域主要包括：计算机视觉、机器学习、医学图像分析。入选斯坦福“全球2%顶尖科学家”、IEEE Senior Member、江苏省“双创博士”、东南大学“至善青年学者”A层次、东南大学“小米青年学者”等。周毅博士已在领域内国际期刊/会议（例如IEEE TPAMI, IEEE TIP, IEEE TMI, CVPR, ICCV, ECCV, ICLR, AAAI, MICCAI等）发表60余篇论文，被引6000余次，8项中/美发明专利，主持国家自然科学基金面上项目、青年项目、江苏省自然科学基金青年项目、中国计算机学会产学研基金等多项项目。 学术兼职包括中国视觉与学习青年学者研讨会（VALSE）执行领域主席，医学图像计算青年研讨会（MICS）执行委员，中国图象图形学学会（CSIG）机器视觉专委会执行委员等。

Brief Bio: Dr. Yi Zhou is an Associate Professor and Doctoral Supervisor at the School of Computer Science and Engineering, Southeast University. His research areas mainly include computer vision, machine learning, and medical image analysis. He has been selected as one of Stanford "Top 2% of Scientists Worldwide", an IEEE Senior Member, a recipient of the Jiangsu Province "Double Innovation Doctoral Program," a Southeast University "Zhishan Young Scholar" (Level A), and a Southeast University "Xiaomi Young Scholar." Dr. Zhou has published over 60 papers in international journals and conferences (such as IEEE TPAMI, IEEE TIP, IEEE TMI, CVPR, ICCV, ECCV, ICLR, AAAI, MICCAI, etc.), which have been cited over 6000 times. He has led several projects, including projects funded by the National Natural Science Foundation of China (NSFC), the Jiangsu Provincial Natural Science Foundation, and the China Computer Federation (CCF) Industry-University-Research Cooperation Fund.

---

袁明志 讲师

taptap点点手机客户端 / Nanjing University of Information Science and Technology

报告题目/Title: 面向精准医疗的人工智能驱动药物发现研究/ AI-Driven Drug Discovery for Precision Medicine

报告摘要：精准医疗旨在通过整合多维数据与智能分析，实现疾病的精准诊断与个体化治疗。其中，药物发现作为精准治疗的重要环节，是推动医疗创新与转化的关键过程。本报告将介绍汇报人近期在构建人工智能驱动的药物发现流程方面的一系列研究工作，涵盖靶点发现、虚拟筛选、ADMET预测等多个阶段。同时，报告还将分享在干湿结合研究方面的探索与实践，展示人工智能在加速药物研发、优化治疗策略以及促进科学与临床融合中的潜力与前景。

Abstract: Precision medicine aims to achieve accurate diagnosis and personalized treatment through the integration of multi-dimensional data and intelligent analysis. As a critical component of precision therapy, drug discovery plays a pivotal role in advancing medical innovation and translational research. This talk will present a series of recent studies on building an AI-driven drug discovery pipeline, covering key stages such as target identification, virtual screening, and ADMET prediction. In addition, the speaker will share explorations of integrating computational and experimental ("dry–wet") research, highlighting the potential of artificial intelligence to accelerate drug development, optimize treatment strategies, and bridge scientific discovery with clinical practice..

个人简介：袁明志，博士，智能医学图像计算江苏高校重点实验室及智慧医疗研究院核心成员。2020年于哈尔滨工业大学获得工学学士学位（通信工程），2025年于复旦大学获工学博士学位（生物医学工程）。主要研究方向为三维计算机视觉、AI4Science（智能药物设计）及智慧病理诊疗等。迄今在ICCV、ECCV、CVPR、ICML、BIB、JBHI等国际顶级会议与期刊上发表论文30余篇，其中第一作者或通讯作者论文18篇。长期担任TPAMI、TIP、RAL及自动化学报等领域重要期刊的审稿人。

Brief Bio: Dr. Mingzhi Yuan is a core member of the Jiangsu Key Laboratory of Intelligent Medical Image Computing and the Institute of Smart Healthcare. He received his B.Eng. degree in Communication Engineering from Harbin Institute of Technology in 2020 and his Ph.D. degree in Biomedical Engineering from Fudan University in 2025. His research interests include 3D computer vision, AI4Science (intelligent drug design), and computational pathology. He has published over 30 papers in top-tier conferences and journals such as ICCV, ECCV, CVPR, ICML, BIB, and JBHI, with 18 papers as first or corresponding author. He also serves as a regular reviewer for leading journals including IEEE TPAMI, IEEE TIP, IEEE RAL, and Acta Automatica Sinica.

---

庄吓海 教授

复旦大学 / Fudan University

报告题目/Title: 可解释人工智能分析/ Explainable Artificial Intelligence Analysis

报告摘要：医学影像智能分析在计算机辅助诊断和治疗等现代医学中发挥着重要的作用；其中，算法的可靠性和安全性对临床应用至关重要。然而，医学影像数据本身的多重特征重叠、跨模态异质、分析算法过程难理解等都对计算方法的复杂性和临床应用的安全可靠性提出了挑战；因此在涉及多模态、跨中心图像和需要弱监督无监督学习的真实场景中，医学图像智能计算方法和模型的泛化能力和可推广性往往不足，难以转化。本次讲座将介绍我们近期提出显式建模和主动构建可解释深度神经网络架构方法，包括基于概念学习的架构可解释和决策可解释算法；实验证明这些通过显式建模获得模型自身可解释性质的方法可以提高其泛化性和可信可靠性。

Abstract: Intelligent medical image analysis plays a crucial role in modern medicine, particularly in computer-aided diagnosis and treatment. The reliability and safety of the underlying algorithms are of paramount importance for clinical applications. However, the inherent challenges of medical imaging—such as overlapping multi-feature characteristics, cross-modal heterogeneity, and the opaque nature of analytical algorithms—significantly increase computational complexity and pose obstacles to clinical safety and reliability. Consequently, in real-world scenarios involving multimodal, multi-center imaging and weakly or unsupervised learning, existing intelligent medical imaging models often suffer from limited generalizability and transferability, hindering clinical translation. This talk will present our recent work on explicit modeling and proactive construction of interpretable deep neural network architectures, including concept-based architecture interpretability and decision interpretability algorithms. Experimental results demonstrate that these methods, which endow models with inherent interpretability through explicit modeling, can significantly enhance their generalization ability, trustworthiness, and reliability.

个人简介：庄吓海，复旦大学教授、博导，大数据学院副院长。研究医学影像信息处理和可解释人工智能分析等。近五年以第一/通讯作者在IEEE TPAMI、TMI等中科院一区期刊发表文章20余篇；多项第一/通讯作者论文获国际组织或顶会奖项，包括2023年获得爱思唯尔出版社-国际MICCAI学会联合颁发的论文最高奖（1st in 1865+，通讯作者），2025年论文获MICCAI 青年科学家奖。入选上海市东方英才计划，获上海市自然科学二等奖（排位第一）和上海市信息学会优秀成果（排位第一）。担任国际MICCAI学会常务理事（全球3位）；担任IEEE TMI、Med Imag Anal等多个中科院一区期刊的编委/副编辑。自2020年来连续入选爱思唯尔-斯坦福大学发布全球顶尖科学家“终身”和“年度”科学影响力榜单。

Brief Bio: Xiahai Zhuang is a Professor and Ph.D. supervisor at Fudan University, currently serving as the Associate Dean of the School of Big Data. His research focuses on medical image information processing and explainable artificial intelligence. Over the past five years, he has published more than 20 papers as the first or corresponding author in top-tier journals, including IEEE TPAMI and TMI. Several of his first/corresponding author papers have received prestigious international awards, including the MICCAI–Elsevier Best Paper Award (1st place out of 1865+, 2023) and the MICCAI Young Scientist Award (2025). Prof. Zhuang has been selected for the Shanghai Oriental Talent Program and has received the Second Prize of the Shanghai Natural Science Award and the Outstanding Achievement Award from the Shanghai Information Society, both as the first-ranked recipient. He serves as an Executive Board Member of the International MICCAI Society (one of only three worldwide) and as an Associate Editor or Editorial Board Member for several top-tier journals, including IEEE TMI and Medical Image Analysis. Since 2020, he has been continuously listed among the World's Top 2% Scientists in both lifetime and annual impact rankings published by Elsevier and Stanford University.

---

周涛 教授

南京理工大学 / Nanjing University of Science and Technology

报告题目/Title:《面向医学影像分析的基础模型及其应用研究》/Fundamental Models Toward Medical Image Analysis and Their Applications

报告摘要：基础模型在自然语言处理等领域的成功，为医学影像分析提供了强大的通用特征与新的研究范式。本报告将首先梳理基础模型的发展脉络与在医学影像中的前沿应用；随后，主要介绍本课题组在基础模型赋能医学图像分割任务上的研究成果；最后，展望并探讨该领域未来的发展方向。

Abstract: The success of foundational models in fields such as natural language processing has provided powerful universal features and new research paradigms for medical image analysis. This talk will first review the development of foundational models and their cutting-edge applications in medical imaging. It will then focus on the research achievements of our team in leveraging foundational models for medical image segmentation tasks. Finally, the talk will provide a forward-looking perspective on the future development directions of this field.

个人简介：周涛，南京理工大学教授、博士生导师，入选国家海外高层次青年人才计划。专注于疾病诊断、医学图像分割、医学图像生成等交叉研究，在领域顶级期刊及会议IEEE TPAMI、IEEE TMI、CVPR、ICCV、MICCAI等发表高水平论文90余篇，谷歌学术引用超9300次，连续入选斯坦福大学发布的全球前2%顶尖科学家年度榜单。担任多个权威期刊（IEEE TIP/TNNLS/TMI/TCSVT）的编委，以及多个顶级会议（AAAI、MICCAI等）AC/SPC。

Brief Bio: Dr. Tao Zhou is a Professor and Ph.D. Supervisor at Nanjing University of Science and Technology, and a recipient of the National "High-Level Foreign Experts Program for Young Talents". His research focuses on disease diagnosis, medical image segmentation, medical image generation, and other interdisciplinary areas. He has published over 90 high-level papers in top journals and conferences such as IEEE TPAMI, IEEE TMI, CVPR, ICCV, and MICCAI, with more than 9,300 citations on Google Scholar. He has been continuously listed on the annual list of the top 2% of the world's leading scientists published by Stanford University. Dr. Zhou serves as an editorial board member for several prestigious journals, including IEEE TIP, TNNLS, TMI, and TCSVT, and is also an AC/SPC for major conferences such as AAAI and MICCAI.