PhD Position in Plant X-ray Tomography and Machine LearningPosted March 19, 2021.
The Faculty of Land and Food Systems (LFS) at the University of British Columbia is located in Vancouver. Doctoral students joining LFS will learn about critical environmental issues in the context of crop production & performance, sustainability, and to our ability to meet our basic human needs (https://www.landfood.ubc.ca/graduate/).
The Canadian Light Source (CLS) is a national research facility located in Saskatoon, Saskatchewan. It is one of the largest science projects in Canada’s history, and a critical tool for Canadian research and development. More than 1,000 academic, government and industry scientists from around the world use us every year, developing innovative solutions in health, agriculture, environment, and advanced materials (https://www.lightsource.ca/).
Mitacs is a national, not-for-profit organization that has designed and delivered research and training programs in Canada for 20 years. Mitacs is committed to its core vision of supporting research-based innovation and continues to work closely with its partners in industry, academia, and government. (https://www.mitacs.ca/en/programs/accelerate). We are inviting application for a joint
Between UBC-LFS and CLS. The successful candidate will be supported by a Mitacs Accelerate Fellowship and their 4-year research project is expected to start in Fall 2021 (September-December). The successful candidate will have the opportunity to contribute their own ideas and to help in finalizing the Mitacs Accelarate Fellowship application (anticipated submission deadline June 30th). The candidate will join the Faculty of Land and Food Systems (UBC-LFS) under supervision of Dr. Thorsten Knipfer and will be co-supervised by Drs. Devin Rippner (USDA-ARS) and Jarvis Stobbs (CLS).
We are looking for a student that is highly motivated to elucidate novel mechanism of how plants cope with water stress by drought. The project will integrate traditional plant water relations with cutting-edge imaging techniques. The student is expected to develop a user-friendly machine-learning tool kit to facilitate high-throughput image analysis and processing. We are seeking a student that has a strong interest in plant biology and agriculture; background with coding and python; experience with machine-learning; and is excited about multidisciplinary research. For more details see ‘Project description’. We offer excellent research facilities at UBC-LFS and CLS (X-ray tomography imaging, machine-learning, image processing) and interdisciplinary scientific training in an international environment.
Qualified candidates holding or expecting to complete their MSc in biology, physics, engineering, or related fields in 2021 are encouraged to apply.
Over the past 10 years, X-ray computed microtomography (microCT) has provided unprecedented insights into structure-function relationships of plants with focus on the impact of drought stress on xylem long-distance water transport, tissue-specific water storage, xylem occlusions, starch storage, leaf porosity, cell shrinkage/expansion, and root cell damage. However, manual image processing following microCT data collection is the big ‘bottleneck’ that makes this technology time-consuming and hinders its broader application for high-throughput analyses of plant performance in response to abiotic and biotic stressors. In this project, we will develop a machine-learning tool kit for automated and high-throughput processing of plant microCT images. Machine-learning tools will provide analytical and computational solutions for 3-D presentation of complex datasets from root, stem to leaf level. Our machine-learning tool kit with be based on an open-source image processing pipeline and will be tailed to (1) elucidate the coupling between vessel embolism formation and water storage throughout the plant, (2) identify drought-induced cell and tissue deformations, and (3) determine recovery from stress through tissue regeneration. We will establish a library of annotated microCT images including six woody species that will serve as training datasets to develop our machine-learning models using a 3d U-net architecture that is based on a convolutional neural network. The ability to more effectively process ‘big microCT imaging data’ will allow us to identify drought resistance traits and physiological performance at speeds that provide Canada’s farmers with new solutions to prepare for rapidly changing climatic conditions.
Applications will be accepted until May 31, 2021 (please email to firstname.lastname@example.org).