A study published in Scientific Reports introduces an artificial intelligence methodology that analyzes three-dimensional dental microwear, the microscopic marks left by food on tooth enamel. This approach aims to reconstruct the diets of fossil primates and hominins with greater consistency, independent of the human analyst.
The technique builds on decades of paleoanthropological research that treats microscopic striations on teeth as archives of dietary history—revealing whether foods were soft or abrasive. Previously, such analysis relied heavily on human interpretation, introducing potential variability.
By using AI to identify 3D wear patterns, the method eliminates subjective bias. The study claims that machine learning can now read these dental archives with repeatability, though specific accuracy rates or comparative performance data were not detailed in the source.
This could transform how scientists reconstruct ancient environments and early human evolution. Reliable dietary inference helps answer fundamental questions about tool use, migration, and social behavior among hominins.
However, the study remains based on modern tooth samples and has not yet been validated on rare fossil specimens, where preservation can degrade microwear features. Broader testing is needed before the methodology becomes a standard tool.