When we are confronted with unclear or vague information about our world, we rely heavily on memory to
fill in missing data and ultimately guide behaviors. For example, if you encounter a massive four-legged animal
alongside its cub while hiking in a forest, your subsequent actions rely not merely on identifying the animal, but
reacting appropriately once you remember that bears are dangerous. What mechanisms in the brain are
responsible for recognizing, identifying, and categorizing objects, all of which can occur within hundreds of
milliseconds? This question is inherently linked to the semantic memory system, which acts as the interface
between incoming sensory information and our preexisting knowledge of the meaning of words, images,
concepts, and their associations. The primary goal of the proposed research is to better understand the
temporal dynamics of how semantic categorization occurs in the human brain.
One method to study semantic categorization is by framing it as a decision that occurs in the brain.
Research on decision-making across several species (e.g., mice, monkeys, and humans) has shown that
decision-making requires a specific computation where evidence is accumulated until a particular threshold is
met, indicating a decision has been reached. This framework (drift-diffusion models, DDMs) has helped to
uncover several decision-making signals across the brain depending on the task, leading researchers to
believe DDMs could be a brain-general mechanism. For this to be true, DDMs should be able to explain activity
in the human ventral visual stream during object recognition tasks. The first aim of this study is to identify
where in the human ventral visual stream can best be described by decision-making signals. The main
hypothesis is that during a semantic categorization task, decision-making signals should occur in the anterior
temporal lobe (ATL), which has been implicated in recent decades as the brain’s semantic hub.
The second aim of this study is to examine how the brain categorizes (e.g., blurred, occluded) visual
stimuli. On a behavioral level, humans require more time to perform object recognition when images are less
clear. One possible explanatory neural mechanism would involve the same neural processes as unambiguous
visual recognition that just occurs more slowly. A contrasting hypothesis is that additional brain areas must be
recruited to solve ambiguity. For instance, accessing stored memories via the medial temporal lobe (MTL) may
be enlisted to call upon previous experiences. Another option relies on a greater degree of cognitive control
from the prefrontal cortex (PFC). This research project will leverage the spatiotemporal precision offered by
intracranial recordings in humans. This study can illuminate cognitive systems and brain networks that are
often damaged in diseases affecting memory, including Alzheimer’s Disease and Semantic Dementia.