Saurabh Mehta

@mehtarg.bsky.social

Janet and Gordon Lankton Professor Founding Director, Cornell Joan Klein Jacobs Center for Precision Nutrition and Health @cpnh.bsky.social Co-Director, PORTENT Director, NIH T32 Artificial Intelligence and Precision Nutrition Cornell Human Ecology

It's been 10 years since we published our first review of sex differences in #TB prevalence. In this update, we found men remain twice as likely as women to have TB, and disparities have likely widened over time. journals.plos.org/plosmedicine...

Peter MacPherson@petermacp.bsky.social · 2mo ago

We updated our review of sex diff. in #TB prevalence In 102 surveys (4.7M participants) men were 2x as likely to have TB than women Evidence consistent with sex differences widening over time 🧪🛟 journals.plos.org/plosmedicine... @nswartwood.bsky.social @kchorton.bsky.social @uofgshw.bsky.social

A map showing countries where surveys were conducted. Bar chart shows the number of surveys per country

Our paper on SciDaSynth: Interactive Structured Data Extraction From Scientific Literature With LLMs was just published in Campbell Collaboration Reviews. It was led by Xingbo Wang and Samantha L. Huey, PhD with Rui Sheng and was a collaboration between our team at @cpnh.bsky.social and Fei Wang.

System workflow of SciDaSynth: (1) Retrieval augmented generation (RAG) based technical framework for extracting and structuring data from figures, text, and tables in scientific documents using LLMs. (2) The user interface then allows for data extraction via question-answering, data validation, correction, summarization, standardization, and database updates through an iterative refinement process.User interface of SciDaSynth. The interfaces features: (A) A query panel for users to input natural language questions or select specific data attributes; (B) A data table displaying extracted information with highlighting of potentially problematic records; (C) Context menu options to validate data by examining relevant document snippets; (D) PDF viewer for accessing original sources; (E) Data standardization panel with multi-level and multi-faceted data summarization and standardization support.

Will this help or hinder human development? Too often we try to predict, based on subtle differences between us and machines; speculating on future trajectories. We argue this is misguided; we should look at the space of complementarity, and shore up the gaps and inequalities between humans.

We look forward to your submissions and please reach out with any questions. Topics may include comparisons of conventional approaches with #PrecisionNutrition, application of AI/ML methods to collection of new data/studies, to existing data, considerations for ethics, and graduate training.

American Society for Nutrition 🍏🔬🌎@nutritionorg.bsky.social · last yr.

#JNutr - CALL FOR PAPERS: #ArtificialIntelligence, #MachineLearning, and #PrecisionNutrition. Editors seek original research, reviews, perspectives & case studies exploring the interface of #nutrition, health, data science & #AI. Deadline: Sep 15, 2025: jn.nutrition.org/tjnut-theme-...

Image to represent Artificial Intelligence, Machine Learning, and Precision Nutrition

Please join us for the inaugural Lindower-Wolitzer DNS Experiential Learning Symposium on May 1, 2025, 2:00 - 3:00pm ET Cornell Ithaca/Zoom. Cornell Human Ecology Public Health and Nutrition Ambassadors will share work on projects relevant to real-world practice. buttondown.com/cornell_pin/...

News from the Program in International Nutrition, Cornell University | Issue #71

We are pleased to share this newsletter with updates and activities in and around the field of international nutrition and global health. Please join us for...

buttondown.com