MRI: Development of a Machine Learning Multimodal Ultrafast Optical Microscope

MRI：机器学习多模态超快光学显微镜的开发

• 批准号: 2117616
• 负责人: Libai Huang
• 金额: $ 76.69万
• 依托单位: Purdue University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-08-01 至 2024-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2117616&HistoricalAwards=false
• 关键词: MRI; Development; Machine; Learning; Multimodal

This award is jointly supported by the Major Research Instrumentation, the Chemical Measurement and Imaging, and the Chemistry Research Instrumentation programs. Purdue University is developing a machine learning multimodal ultrafast nonlinear optical microscope to support the research of Professor Libai Huang and colleagues Gregery Buzzard, Sujith Puthiyaveetil, Michael Reppert, and Chi Zhang. In general, this instrument development combines expertise in instrument design, non-linear ultrafast spectroscopy, microscopy, and machine learning. If successful, the resulting instrument will represent an enabling tool that could facilitate investigations involving complex materials and biological systems over a wide range of time (10 femtoseconds - microseconds) and length (50 nm - micron) scales, extending capabilities beyond what is currently available with conventional commercial microscopy instruments. This temporal/spatial information may be used to better understand energy and heat flow in complex materials and biological samples. This instrument will enhance education, research, and teaching efforts of students at all levels, in several departments, as well as be accessible for use at other institutions. The award to develop a machine learning multi-modal ultrafast optical imaging platform is aimed at enhancing research and education at all levels, especially in areas such as optical microscopy, machine learning, and ultrafast spectroscopy by reducing optical exposure and measurement time by about 100-fold without significant loss in reconstructed image quality. Studies focused on coherent and non-equilibrium energy transport in nanomaterials, multi-scale tracking of light response in photosynthetic membranes, and heat flow in biological assemblies are to be pursued as are those focused on machine learning-enabled adaptive sampling for ultrafast microscopy measurements. This instrument development project has the promise of opening up optical imaging studies of complex materials or biological systems at time and length scales beyond what is currently available with conventional commercial optical microscopy instrumentation.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

该奖项由主要研究仪器、化学测量和成像以及化学研究仪器计划共同支持。普渡大学正在开发一种机器学习多模超快非线性光学显微镜，以支持黄立白教授及其同事Gregery Buzzard、Sujith Puthiyaveetil、Michael Reppert和张驰的研究。总的来说，这种仪器的开发结合了仪器设计、非线性超快光谱、显微镜和机器学习方面的专业知识。如果成功，所产生的仪器将是一个有利的工具，可以促进涉及复杂材料和生物系统的大范围时间(10飞秒-微秒)和长度(50纳米-微米)的调查，将能力扩展到目前传统商业显微镜仪器所具有的能力之外。这种时间/空间信息可以用来更好地理解复杂材料和生物样品中的能量和热流。这一工具将加强各级学生在几个系的教育、研究和教学努力，并可在其他机构使用。开发机器学习多模式超快光学成像平台的奖项旨在加强所有级别的研究和教育，特别是在光学显微镜、机器学习和超快光谱等领域，通过将光学曝光和测量时间减少约100倍，而不会对重建图像质量造成重大损失。将继续进行纳米材料中相干和非平衡能量传输、光合膜中光响应的多尺度跟踪以及生物组件中的热流的研究，以及那些专注于超快显微镜测量的机器学习自适应采样的研究。这一仪器开发项目有望开启复杂材料或生物系统的光学成像研究，其时间和长度超过目前传统商业光学显微镜仪器的可用范围。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。