Proof‑of‑concept for using machine learning to facilitate data extraction for human health chemical assessments: a study protocol

‍Title: Proof‑of‑concept for using machine learning to facilitate data extraction for human health chemical assessments: a study protocol‍

Evidence-Based Toxicology Journal

Authors: Michelle Angrish, Kristina A. Thayer, Brittany Schulz, Artur Nowak, Amanda Persad, Allison L. Phillips, Glenn Rice, Teresa Shannon, A. Amina Wilkins, Krista Christensen, Elizabeth G. Radke, Andrew Shapiro, Michele M. Taylor, Vickie R. Walker, Andrew A. Rooney & Sean M. Watford

What is the report about?

This protocol's objective is to present the methodology that aims to discover whether a data extraction tool, Dextr—Laser AI’s predecessor, could improve user experience and the chemical assessment process. The objective stems from the need to improve traditional approaches to the chemical assessment workflow, which cannot be adapted.

How did Laser AI help?

Based on the protocol methodology, the research team will test machine learning capabilities in the chemical assessment workflow by using the Environmental Protection Agency (EPA) 's systematic evidence map (SEM) and the semi-automated data extraction tool Dextr (Laser AI).

Citation: Angrish, M., Thayer, K. A., Schulz, B., Nowak, A., Persad, A., Phillips, A. L., … Watford, S. M. (2024). Proof‑of‑concept for using machine learning to facilitate data extraction for human health chemical assessments: a study protocol. Evidence-Based Toxicology, 2(1). https://doi.org/10.1080/2833373X.2024.2421192