How do we encode the details of our memories?

---

Photo by Ala on Unsplash

By Ainsley Kohler, BC '23
Posted on September 14, 2023

Visualize the setting you ate breakfast in this morning. You will likely find that you can picture objects that were around you, what color they were, where they sat, and how they were positioned. You might even be able to recall that, as you ate breakfast this morning, you listened to the sound of the dishwasher running. This ability to recall a plethora of details associated with a context is a key feature of episodic memory.

Our memories can be broadly divided into two subgroups: those that last only up to a few minutes, including our sensory, short-term, and working memory, and those which can last days, months, years, and even a lifetime, which we refer to as our long-term memory. Within long-term memory, we can make further distinctions, including differentiating between memories that do not require conscious recall such as procedural memories (think about the last time you rode a bike: did you have to think about how to do it, or does it just seem to come naturally?) and memories that we conscious recall. We call the latter category declarative memories, which can also be separated into two subgroups: semantic and episodic memory. The former consists mainly of factual information, while episodic memory holds all of our specific personal experiences from a particular time or place.

Our episodic memories, like those of our last meal, are rich in detail, but how do we accomplish this? What are the neural mechanisms underlying the binding of certain aspects of a scene to their context? Cooper et al. 2020 is a study that looks at regions of the brain that process different features of memory and asks whether there are areas that might be tied to feature binding, or the integration of details into one memory of an event. It might be hard to imagine how we could recreate detail-rich episodic memories in the lab that would allow us to effectively explore these topics. Researchers must be able to present participants with a scene that has multiple, distinct details which can be effectively recalled later in a memory test. In Cooper et al. 2020, participants start by viewing an item in a context that they are told to remember. The item has a unique color, background, and is paired with an audio track that is either negative or neutral. Later, participants will be asked to recall the three aspects of the scene. During recall, they are presented with greyscale versions of objects they saw again and asked to do three things: First, they rotate the panoramic background behind the object until they believe they have lined up the item to the position it was presented in. Second, they use a color wheel to manipulate the color of the grey object until they believe it matches the color it was during encoding. Finally, they identify whether the sound associated with the object was negative or neutral.

In this way, the study looked at three aspects of a memory that are combined together in one event: location (1), color (2), and sound (3).

There are brain regions that are associated with different scene features: object color is represented in the V4, or the visual cortex, in the inferior occipital lobe, while location is often found in the parahippocampal place area (PPA) in the inferior tempo-occipital cortex, and sound processing is associated with the primary auditory cortex in the temporal lobe. Cooper et al. 2020 verified that brain activity in these areas is predictive of the ability to later recall the details they are associated with. However, the study goes beyond these areas and looks for regions of the brain that are active when participants are viewing a scene that are predictors of remembering not just one detail, but multiple details together. Cooper et al. 2020 is able to accomplish this by starting with the results of the memory test. The study asks what the difference in brain activity is between trials where participants remembered more than one detail correctly. Through this method of working backward, areas of the brain that are associated with feature binding, or the successful integration of multiple details of a memory, can be identified.

The region of the brain that Cooper et al. 2020 focuses on after running these analyses is the left hippocampus. As shown below, we can see that the activity in the left hippocampus was greater during trials where the participant would later successfully recall color and scene, scene and sound, and all three together. Interestingly, there is less activity in the left hippocampus in trials where participants only recalled color and sound, but not scene. This indicates that the left hippocampus may play a role in the binding of scene (or location) information to other details of a memory, including color and sounds. Another distinct feature of the left hippocampal activity that helped researchers to conclude it may be associated with feature binding was that occurred later in the encoding phase. Participants viewed the objects in their contexts for 6 seconds, which may not seem like a long time, but within that 6 seconds, there are many changes in brain activity. While earlier activity in this 6-second window is in areas that help the participants perceive features of an event, later activity, like that in the hippocampus, might be associated with integrating just viewed details together into one event. The data collected supported this hypothesis: activity in the hippocampus early in encoding is not a great predictor of the number of details recalled, but activity at the end of encoding is.

So now we know one of the regions that might help us to recall all the little details of our breakfast this morning: the color of the food, the sound of the dishwasher running, and where the plate sat on the table. While you were eating this morning not only was your brain taking in each feature of the scene and helping you make sense of what you were perceiving, but also binding features together so that you recall them as one event: breakfast. Going forward from Cooper et al. 2020, future studies might choose to look more closely at the timing of these different processes that occur to ensure that they are truly a product of separate encoding events and not regional differences in blood flow. We can be sure, though, that the process of making memories is a multifaceted one that relies on feature binding to create cohesive events.

---