Postdoctoral Research Associate Applied Machine Learning in X-ray Imaging

Website BrookhavenLab National Synchrotron Light Source (NSLS-II)

Job ID 1969
Brookhaven National Laboratory is a multipurpose research institution funded primarily by the U.S. Department of Energy’s Office of Science. Located on the center of Long Island, New York, Brookhaven Lab brings world-class facilities and expertise to the most exciting and important questions in basic and applied science—from the birth of our universe to the sustainable energy technology of tomorrow. We operate cutting-edge large-scale facilities for studies in physics, chemistry, biology, medicine, applied science, and a wide range of advanced technologies. The Laboratory’s almost 3,000 scientists, engineers, and support staff are joined each year by more than 4,000 visiting researchers from around the world. Our award-winning history, including seven Nobel Prizes, stretches back to 1947, and we continue to unravel mysteries from the nanoscale to the cosmic scale, and everything in between. Brookhaven is operated and managed by Brookhaven Science Associates, which was founded by the Research Foundation for the State University of New York on behalf of Stony Brook University, and Battelle, a nonprofit applied science and technology organization.


We have a lively, fast growing data science research program at BNL, with a specific focus on the challenges presented by the analysis, interpretation, and use of data at extreme scales and in real time. The data science program is accompanied by significant computational modeling research effort, in support of the design, planning, analysis, and interpretation of experiments and results. Currently, two departments of NSLS-II (National Synchrotron Light Source II) and CSI (Computational Science Initiative) have a joint effort and plays the key role in the data science research.

NSLS-II is a state-of-the-art synchrotron facility enabling the study of material properties and functions with nanoscale resolution and exquisite sensitivity by providing world-leading capabilities for X-ray imaging and high-resolution energy analysis. The Imaging and Microscopy Group at the NSLS-II operates 5 state-of-the-art imaging beamlines (FXI, HXN, SRX, XFM, TES). In particular, the Full-Field X-ray Imaging (FXI) beamline has the word leading full-field nanoscale imaging capability in the world. The CSI provides a laboratory-wide umbrella for these activities, bringing together computer scientists, applied mathematicians, and domain scientists to carry out leading edge research, convert research results into practical solutions that advance domain science, and provide the necessary computing infrastructure services and training to support efficient operation.

NSLS-II and CSI at BNL is seeking an exceptional postdoctoral researcher to develop machine learning algorithms to enhance the sensitivity and resolution of X-ray imaging at NSLS-II.

The position is funded through a LDRD (Laboratory Directed Research Development) project at BNL. The candidate will work closely with NSLS-II’s Imaging and Microscopy program (IMP) and CSI, with a focus on the development of machine learning based imaging processing tools to denoise scientific high-dimensional images in order to achieve scientific discoveries. Generally, scientific images are highly correlated, and the intensity of those images is governed by the physics imposed by material properties. We are expected to incorporate material physics into machine learning models to recover noisy images and achieve self-consistence, with a goal of enhancing the image resolution and detection sensitivity. Successful execution of this project will not only enhance the imaging capability of the FXI beamline but will lead to opportunity to establish new methodology for the world-wide imaging community. The successful candidate will collaborate with a diverse team of scientists to collect data, analyze data, refine models and publish outcomes.

Essential Duties and Responsibilities:

  • Develop advanced imaging analysis method dedicated for recovering noisy data using machine learning
  • Work with experts in the field of X-ray imaging and computer science to optimize the developed method.
  • Conduct experimental work to verify the developed method.
  • Report research results in peer-reviewed journals and at conferences.

Required Knowledge, Skills, and Abilities:

  • PhD degree in Computer Science, Physical Sciences, Applied Mathematics or related field.
  • Demonstrated experience in imaging processing
  • Experience in applied machine learning framework and algorithm development
  • Experience of working in Linux environment
  • Proficient in python language for scientific data analysis

Preferred Knowledge, Skills, and Abilities:

  • Proficient in machine learning (neural network) framework such as PyTorch or Tensorflow.
  • Experience in customizing machine learning code and implementation
  • Experience of image processing algorithms development
  • In-depth knowledge in Python or C++
  • Experience with high performance computing systems, including accelerators (GPU).

Other information: BNL policy requires that research associate appointments be made to individuals who have received their doctorate within the past five years.