Nutrition in the West is a young discipline, only 100 years old. Artificial intelligence is even newer. Combining the two could create a powerful tool for customizing health advice rather than the current one-size-fits-all approach.
“By harnessing the power of AI, we can now efficiently personalize nutritional recommendations by taking into account the biology of an individual or groups of people and, importantly, the food environment in which they live. We can also contextualize the advice to suit,” said Dr. Saurabh Mehta. , Professor Janet Lankton and Gordon Lankton in the Department of Nutritional Sciences, School of Human Ecology.
Mehta’s team uses generative AI to power health dashboards and chatbots that use only clinical information and data from high-quality, evidence-based sources. It is one of his dozens of ambitious, interdisciplinary projects at Cornell University, powered by Empire AI, his $400 million effort to build a shared academic research computing facility. will be done. Users of its facilities, including Cornell faculty and students, aim to advance responsible research and AI opportunities focused on the public good.
The consortium, announced by Governor Kathy Hochul on April 22, is comprised of seven major New York universities and research institutions, including Cornell University.
“Research creating and using AI is already changing the world, and Cornell researchers are leading the way in developing responsible and trustworthy AI tools for maximum benefit.” Vice President Kristin Van Vleet said.
“We are very grateful to our state and other contributors. Empire AI is a significant investment in a shared academic research computing facility that will impact fields from agriculture to healthcare to urban development.” she stated. “Cornell University and a consortium of other great New York State universities and research institutions are partnering on a new way to imagine this, and now we are making it a reality.”
“Fundamental research in generative AI requires vast amounts of computing power that are not currently available to academics,” said Cornell University Ann S. Bowers Dean of the College of Computing and Information Sciences and lead dean of the Cornell AI Initiative. says Kavita Bala. “Empire AI will change this and enable core research in generative AI models in academia that can be applied to the benefit of society in a wide range of fields including nutrition, climate change, materials discovery, health and medicine, and physics. ”
Cornell University researchers are already piloting and laying the groundwork for research efforts that leverage AI to benefit the daily lives and health outcomes of New Yorkers. Mehta, founding director of the Cornell Center for Precision Nutrition and Health (CPNH), and CPNH researcher Samantha Huey are using AI in a variety of applications to help people make nutrition and health decisions for end users. I am making use of it. This means not only consumers, but also those in government and non-government organizations who are focused on establishing and following policy effectively and efficiently.
Mehta predicts that AI will accelerate the implementation of other nutrition and food-to-medicine programs, such as medically tailored diets and food prescriptions.
Mehta, who is affiliated with the World Health Organization’s Office of Nutrition and Food Safety, said, “AI should influence the dissemination and dissemination of dietary guidelines for Americans, and help define policy and quantify subsidies.” (CPNH hosts a WHO Collaborating Center). Cornell University’s Nutrition for Health Research) and Weill Cornell Professor of Population Health Sciences Fei Wang and others were involved in the study. CPNH is also home to the Artificial Intelligence and Precision Nutrition Training Program, the first National Institutes of Health-supported training grant focused on AI.
Empire AI’s founding institutions (also including Columbia University, New York University, Rensselaer Polytechnic Institute, State University of New York, City University of New York, and the Flatiron Institute) will use the funding to advance computational research, including data analysis. Masu. , theory, modeling and simulation.
Greg Morissette, the Jack and Lila Kneafsey President and Vice Chancellor of Cornell Tech, said this type of funding allows Cornell and other institutions to achieve more than any university could alone. He said it will be possible to work on large-scale models.
“Right now, the United States is a major supplier of AI models,” he said. “But if we want to keep pace with other countries and states like California and Massachusetts, this type of state-level investment to support research is essential.”
At Cornell University, artificial intelligence models are being used in many other medical settings.
Rashit Saluja, a doctoral student in the Department of Radiology at Weill Cornell Medicine, is using this new technology to codify the language radiologists use to describe the images they read. It is systematized. This is an advancement aimed at reducing the burden on physicians and streamlining patient care. Saluja is collaborating with Mart Sabunk, a professor at Cornell Polytechnic Institute and Cornell School of Engineering and his associate director for AI and engineering research in the Weill Cornell University Department of Radiology. Currently, Saluja and his team are collecting multimodality datasets and looking at how their models can incorporate it all, with the goal of creating a text suggestion system.
Saluja says radiologists’ reports contain large amounts of unstructured text data, with little common nomenclature to describe what they’re seeing in the images. “This text data is so valuable that no AI has been able to analyze it before. Once the image is in the system, the model examines it and makes recommendations to the radiologist.”
He said the AI model can provide some recommendations about the text to avoid biasing radiologists.
“Radiologists in the United States have a busy workload,” Saluja said. “If we can develop smart workflows, they will become more efficient and allow us to make better decisions. The goal is not to replace radiologists.”
Saluja said the research will take about a year and a half to implement, depending on resources, and that these AI models could be developed separately for different areas of radiology, such as musculoskeletal, neuroradiology, and mammography. He said that it could also be developed as a general-purpose model. It can be fine-tuned for specific applications.
For patients, improved healthcare efficiency means shorter wait times and, in some cases, more accurate diagnoses. For shoppers, increased food production efficiency means more money in their hands. Akshay Ajagekar, a systems engineering doctoral student at Cornell University, has prototyped an AI-based controller that he says will reduce energy use by 30% in lettuce greenhouses across the state. This is an innovation that has the potential to lower retail produce prices for New York shoppers.
Source: cornell.edu
