In collaboration with at Beaumont, Jing Zhang, associate professor of computer science, has received funding from the National Institute of Food and Agriculture, a federal agency within the United States Department of Agriculture, for a three-year research project titled “A Digital Rice Selection System that Integrates UAV Imaging, Machine Learning, and Multi-Trait Decision-Making” (USDA-NIFA Award #2022-67021-36584, $650,000, Lead PI: Yubin Yang, Texas A&M AgriLife Research Center at Beaumont. 海角社区 receives $160,000 as the sub awardee, Lamar’s PI: Jing Zhang).
The primary goal of the project is to transform traditional rice crop research toward a digital and technology-centered research paradigm. The project will develop advanced machine learning algorithms to extract key rice phenological, morphological, and architectural traits from UAV imagery during critical rice growth stages, create multi-trait-based machine learning models for final aboveground biomass and rice grain yield estimation and build a multi-criteria decision-making system for best-performing rice genotype selection. Zhang and his research team at 海角社区 will help with the development of machine learning algorithms for crop feature extraction from UAV images acquired at different crop stages, the integration of multi-dimensional and multi-temporal traits and the implementation of the digital rice selection system.
“My research collaboration with the Texas A&M AgriLife Research Center at Beaumont started in 2018. This grant will greatly strengthen this institutional-level collaboration and provide me a good opportunity to apply advanced computer science techniques to solve cutting-edge problems in the agriculture area. I believe we will have exciting research findings and top-level publications. In addition, this project will enhance teaching and student professional development at LU," he said. "The new teaching materials based on the research outcomes will be integrated into my Image Processing and Computer Vision courses to enrich curricula, and the selected computer science students will participate in the project to obtain hands-on research experience, boost their research abilities and increase their competitiveness for future careers.”
Zhang also has been awarded a patent titled “Systems and Methods for Extracting Prognostic Image Features” (Invertors: Maciej Mazurowski, Duke University; Lars Johannes Grimm, Duke University; Jing Zhang, 海角社区). This patent is useful for determining lesion severity and disease prognosis for breast cancer, which is a leading cancer diagnosis among women with more than 230,000 new cases a year in the United States. The patented systems and methods can extract and analyze image features that are correlated with breast cancer molecular subtypes and prognosis and, therefore, provide clinical benefit for breast cancer by identifying cancer subtypes without formal genetic analysis.
This patent application was filed by Duke University in 2016 and was granted in 2021. The patent presents methods and systems to analyze breast cancer molecular subtypes using dynamic imaging features extracted by advanced medical image analysis algorithms. Breast cancer can be classified into four distinct molecular subtypes: luminal A, luminal B, human epidermal growth factor receptor 2 (HER2) enriched and basal-like, based on the differences between tumors at the genetic level. These distinct differences in tumor genetics cause different patterns of disease expression, response to therapy, and patient survival outcomes, so breast cancer subtype analysis is very important. These patented methods and systems discovered that new breast MR dynamic features, which compare the average breast cancer lesion image enhancement to a maximal breast cancer lesion image enhancement at a time point when a volume of a background parenchyma image enhancement reaches a predetermined threshold, are correlated with breast cancer subtype and are useful for determining cancer lesion severity and disease prognosis.
“I am very excited about this patent and believe it will benefit breast cancer patients by providing an efficient method for breast cancer subtype identification with minimal additional cost, time, or infrastructure investment,” Zhang shared.