基于该项全球首创技术，团队在国际上首次实现单神经元水平"输入—输出"的准确链接，为深入解析脑 功能的神经环路机制提供了新的机遇。结合团队自主研发的重建算法，VIVIT技术还可构建从亚细胞到 全器官的三维图谱，精准再现包括细胞器、神经突触在内的精细结构和位置，为各种器官绘制纳米级三 维立体"画像"。 （文章来源：科技日报） 为了兼顾透明和无形变，团队开展跨学科研究，自主研制出高折射率的离子液体。"基于该离子液体， 我们开发了整套VIVIT组织透明化处理技术，首次将不透光的生物组织在低温下转变为'玻璃态'。"苑克 鑫介绍，玻璃态的样本不会由于形成冰晶对组织造成机械性损伤，可在-80℃下长期保存，随用随取。 论文中展示了啮齿类、非人灵长类和人类脑组织经处理后，神经突触等亚细胞级精细结构在显微镜下均 清晰可见。团队还发现，离子液体处理后的样本经荧光染料染色后，其信号强度可提升2—30倍。这为 跨尺度进行三维数据获取与空间重建奠定了坚实基础。 11日，清华大学生物医学工程学院苑克鑫教授团队全球首创的生物组织透明化处理方法登上国际期刊 《细胞》杂志。该方法能将不透明的生物组织转变为"玻璃态"，便于显微镜等设备对组织内部进 ...

图像复原领域长期存在一个两难困境：追求高质量往往耗时长，而追求高效率则难免损失细节精 度。如何把一张老照片修复得又快又好？ 近日，中国科学院深圳先进技术研究院发布HYPIR图像复原大模型。它由该院先进计算与数字工程 研究所研究员董超团队研发，为图像复原技术的实际应用提供了更高效的解决方案，更为影视修复、文 化传承与保护等领域带来新的可能。 传统方法中，基于预训练扩散模型的复原技术显著提升了图像复原效果，但存在计算复杂度高、推 理速度慢、训练资源消耗大以及生成结果可控性不足等问题。这些成为限制图像复原技术发展的瓶颈。 2024年，董超团队研发出智能画质增强大模型SUPIR。它能将低质量的图像恢复到接近原始状态的 高清图像，有效修复多种退化类型的图像。 HYPIR作为SUPIR的升级版，舍弃了迭代式扩散模型训练，改用单步的对抗生成模型训练方式，将 原有的算法速度提升了数倍。同时，HYPIR采用更新的文生图基模型进一步提升算法效果，实现8K级 别的细节生成，在生成图像的稳定性和可控性方面远超SUPIR大模型。 "以往的图像复原往往包括扩散模型蒸馏等过程。HYPIR则不需要依赖这些步骤，复原方法更加简 单。它在训练和 ...

针对上述难题，研究团队在人造细胞内实现了核苷酸从头合成代谢通路的构建，并将合成的核苷酸 用于RNA转录。该代谢通路以碳酸氢铵为起点合成尿苷三磷酸（UTP），包括氨甲酰磷酸合成酶 （CPS）、天冬氨酸转氨甲酰基酶（ATC）、二氢乳清酶（DHO）、二氢乳清酸脱氢酶（DHODH）、 磷酸核糖焦磷酸激酶（RPPK）、尿苷5′-单磷酸合成酶（UMPs）、尿苷酸激酶（UK）和核苷二磷酸激 酶（NDK）8种酶。同时，研究引入由肌酸激酶（CK）和磷酸肌酸组成的ATP（三磷酸腺苷）再生系统 以驱动该代谢途径。在优化条件下，3小时内可产生0.85毫摩尔每升的UTP，与CTP、GTP、ATP和T7 RNA聚合酶共同实现了人造细胞内的RNA转录。人造细胞内UTP从头合成代谢通路及其后续RNA转录 的成功实现，为自主人造细胞的构建奠定了坚实基础。 核苷酸是合成RNA的必要成分。以简单化合物从头合成核苷酸，对生命活动来说至关重要。目 前，人造细胞研究中转录过程均依赖预先添加的核苷酸，而非内源合成，这限制了人造细胞的自主性和 长期稳定性，在人造细胞内原位合成核苷酸以维持RNA转录仍是巨大挑战。 科技日报讯 （记者朱虹 通讯员梁英爽）8月8 ...

这些癌细胞仿佛在等待一个信号，而病毒引发的炎症，正是那个"启动密码"。美国科罗拉多大学安 舒茨医学分校的科学家在小鼠模型上做了实验：当乳腺癌小鼠呼吸道感染后，原本平静的肺部迅速变 得"热闹"起来，休眠的癌细胞在几天内开始苏醒、分裂，短短两周，转移病灶就明显扩大。 科学家们紧接着分析了英国生物样本库（4837人，所有癌症类型）和美国Flatiron健康数据库 （36845人，乳腺癌患者），发现人类世界的数据也有类似的警报：呼吸道感染后的癌症患者，其癌症 相关死亡风险是未感染者的一倍；而在乳腺癌患者中，感染后肺部转移的风险上升了40%以上。 这意味着，呼吸道感染可能"开门"让癌细胞趁虚而入。或许，人体免疫系统在全力抗敌时，无意中 为癌细胞创造了"温床"，炎症信号就像"营养液"，刺激它们重新活跃。 研究无止境，科学无定论。这一研究还有待进一步扩大样本和深入分析。它也不是让人们恐慌的理 由，而是提醒人们：健康是一场精密的平衡游戏。对于癌症康复者来说，预防呼吸道感染，比以往任何 时候都更重要。戴口罩、打疫苗、勤洗手，这些看似简单的动作，可能就是在为自己的身体筑起一 道"防火墙"。 《自然》杂志最近发表了一项重磅研究，揭 ...

Group 1: Lunar Exploration Achievements - The Chang'e 6 mission has successfully unveiled the evolutionary history of the moon's far side, marking a significant milestone in lunar exploration [1][5] - The mission collected and returned samples from the moon's far side, achieving multiple breakthroughs in key technologies [6][5] - The Chang'e 6 mission returned 1935.3 grams of lunar samples, setting several world records during its 53-day mission [4][6] Group 2: Human Spaceflight Developments - The Chinese space station has been constructed with all components being domestically produced, achieving 100% self-sufficiency in key technologies [3][5] - The space station has hosted nine crewed missions and conducted six "space rendezvous" events, showcasing China's advancements in human spaceflight [3][9] - The upcoming missions include the development and launch of additional modules to enhance the capabilities of the Chinese space station [3][10] Group 3: Deep Space Exploration Initiatives - The Tianwen-1 mission successfully achieved three major objectives: orbiting, landing, and roving on Mars, marking a first in global space exploration [9] - Future missions include Tianwen-3 for Mars sample return and Tianwen-4 for Jupiter system exploration, indicating China's commitment to deep space exploration [10] - The National Space Administration is also focusing on solar exploration and the development of a new generation of civil space infrastructure [10]

Group 1: Lunar Exploration Achievements - The Chang'e 6 mission has successfully returned samples from the far side of the moon, marking a historic first in human history [1][5] - The mission lasted 53 days and set multiple world records, with a total of 1935.3 grams of lunar samples collected [5][6] - Key technological breakthroughs were achieved during the mission, including advancements in lunar reverse orbit design and intelligent rapid sampling [6] Group 2: Manned Spaceflight Developments - The construction of the Chinese space station began with the successful launch of the Tianhe core module on April 29, 2021, leading to a series of successful missions [3] - The Chinese space station has completed its basic configuration in orbit and achieved 100% domestic production of key components [3] - The space station has hosted nine crewed missions and conducted six "space rendezvous" operations, showcasing China's rapid advancements in manned spaceflight [3][9] Group 3: Future Exploration Plans - The China National Space Administration (CNSA) is accelerating the implementation of the Chang'e 7 and Chang'e 8 missions, aiming to establish an international lunar research station [6][10] - Upcoming deep space exploration missions include the Tianwen-3 Mars sample return and Tianwen-4 Jupiter system exploration [10] - The CNSA plans to develop a new generation of national civil space infrastructure and satellite internet systems to enhance global contributions [10]

Core Viewpoint - The research conducted by the University of Science and Technology of China, in collaboration with international partners, successfully expands the parameter space for direct detection of strongly interacting dark matter using data from the James Webb Space Telescope, overcoming limitations of ground experiments [1][2] Group 1: Research Findings - The study reveals that detecting the nature of dark matter is crucial for enhancing the particle physics standard model and understanding the formation and evolution of the universe [1] - Ground experiments have blind spots for detecting strongly interacting dark matter, which this research addresses by utilizing the near-infrared spectrometer's dark images from the James Webb Space Telescope [1] - A new image processing method was proposed to effectively eliminate high-energy event interference, resulting in new constraints on dark matter detection [1] Group 2: Implications and Future Applications - The research highlights the unique advantages and feasibility of space-based detectors for strongly interacting dark matter [2] - The analysis methods developed can be applied to data from other space telescopes, providing a solid theoretical foundation for the design and data processing of future space dark matter detectors [2] - This work is expected to play a significant role in the field of direct dark matter detection [2]

图像复原领域长期存在一个两难困境：追求高质量往往耗时长，而追求高效率则难免损失细节精度。如 何把一张老照片修复得又快又好？ 近日，中国科学院深圳先进技术研究院发布HYPIR图像复原大模型。它由该院先进计算与数字工程研究 所研究员董超团队研发，为图像复原技术的实际应用提供了更高效的解决方案，更为影视修复、文化传 承与保护等领域带来新的可能。 传统方法中，基于预训练扩散模型的复原技术显著提升了图像复原效果，但存在计算复杂度高、推理速 度慢、训练资源消耗大以及生成结果可控性不足等问题。这些成为限制图像复原技术发展的瓶颈。 2024年，董超团队研发出智能画质增强大模型SUPIR。它能将低质量的图像恢复到接近原始状态的高清 图像，有效修复多种退化类型的图像。 HYPIR作为SUPIR的升级版，舍弃了迭代式扩散模型训练，改用单步的对抗生成模型训练方式，将原有 的算法速度提升了数倍。同时，HYPIR采用更新的文生图基模型进一步提升算法效果，实现8K级别的 细节生成，在生成图像的稳定性和可控性方面远超SUPIR大模型。 "以往的图像复原往往包括扩散模型蒸馏等过程。HYPIR则不需要依赖这些步骤，复原方法更加简单。 它在训练和 ...

Group 1: Industry Overview - The 2025 SIA Shanghai International Smart Factory Exhibition showcased over 1,000 companies presenting various collaborative robots and intelligent control systems, emphasizing the shift from traditional manufacturing to intelligent manufacturing [1] - China has become the world's largest robot producer, with industrial robot production increasing from 33,000 units in 2015 to 556,000 units in 2024, maintaining the largest industrial robot market for 12 consecutive years [2] - The integration of AI technology into industrial robots is a significant trend, enabling them to autonomously perceive, analyze, and make decisions in complex environments [2] Group 2: Technological Advancements - The intelligent pipeline polishing robot developed by Shanghai ZaiKe Intelligent Technology Co., Ltd. can operate in confined spaces, improving work efficiency by over 10 times compared to manual labor [2] - The smart composite robot from Foshan Zhiqing Technology Co., Ltd. features a dual-brain architecture for precise and smooth operations, capable of working in temperatures from 0°C to 50°C with a positioning accuracy of ±0.05 mm [3] - Electric grippers, which are essential for automation, currently have a low market penetration of only 6% in China, indicating significant growth potential [4][5] Group 3: Vertical Integration and Collaboration - Companies are pursuing vertical integration across the entire value chain to provide customized services for industries such as power batteries, automotive, electronics, and semiconductors [4] - A collaboration between Yikaida Technology (Shanghai) Co., Ltd. and Huayan Robotics (Jiangsu) Co., Ltd. has led to standardized solutions in machine tool loading and spraying, enhancing deployment speed for downstream clients [6] - The China Robot Industry Alliance predicts that by 2025, industrial robot sales in China will exceed 400,000 units, with the market size expected to reach 150 billion yuan [6] Group 4: Expanding Applications - Robots are extending their applications beyond industrial sectors into areas like healthcare, with the introduction of acupuncture robots that utilize 3D vision sensors for precise point location [7] - The acupuncture robot employs graphene purification technology to eliminate smoke generated during the process, making it more suitable for patients with respiratory issues [7] - Companies are also focusing on collaborative robots to enhance services in welding, polishing, and material handling within industrial settings [7] Group 5: Market Dynamics - The exhibition highlighted the robust growth momentum of China's robot market, driven by policy support, technological breakthroughs, and increasing demand across various industries [8]

记者8月7日从苏州工业园区获悉，由苏州工业园区洛加大先进技术研究院（以下简称"洛加大苏研院"） 牵头打造的冷冻电镜与AI双引擎平台日前面向全球发布。该平台整合了冷冻电镜前沿技术与海量数 据，有助于生物医学从基础研究到临床应用的全方位发展。 （文章来源：科技日报） 此次发布的冷冻电镜与AI双引擎平台，涵盖自研的透射电镜相机、冷冻电镜、SBDD/AIDD生物医学计 算平台、医疗大数据智能辅助诊疗与科研平台等软硬件协同创新产品。 其中，"白云4M/16M"透射电镜相机等国产化产品由洛加大苏研院科研团队联合苏州德锐特成像技术有 限公司联合研发。该公司是全球领先、国内首家电子探测器企业，也是该领域国内唯一实现量产的企 业。 冷冻电镜和SBDD/AIDD生物医学计算平台立足电镜成像，向更多显微领域拓展，让AI能看见并理解微 观世界。其中，200千伏冷冻电镜是苏州地区首台可提供冷冻电镜服务的设备，它搭载高性能探测器， 能提供原子级分辨率成像服务，可应用于药物递送机制研究、蛋白质结构解析等多个领域。 SBDD/AIDD生物医学计算平台融合结构药物设计与人工智能药物发现技术，通过强大算法精准筛选药 物靶点、设计先导化合物，将大 ...