Memory creation part-1

November 24, 2018

Memory creation part-1

(Photo : Google pics)

Hello Everyone, I am back with my Article and today its time to generate a memory with similar data a person have. Interesting isn't it...!! In fact, we both use a similar mind map as shown below.

(Photo : Google pics)

Human Nature

(Photo : Google pics)

A.I. Mind Map

Let's go on with the Human brain first. Imagine you have just come back from a fantastic holiday, in which you explored a far-off part of the world and saw incredible landscapes and rare animals. The many pictures you took cannot fully convey the way you remember these events. Wouldn't it be nice if your friends could directly ‘read’ the content of your memories of that trip? Well, maybe this idea is not as crazy as it seems.

Progress in neuroscience has made it possible to associate perceptual or cognitive processes with particular brain structures. However, because a given brain region is often associated with several processes, a person's mental state or perceptual experience can only rarely be determined from the average activity of a single brain area.

Recently, however, approaches that analyse the signal from many volume elements (voxels) of the brain simultaneously have started to change the game. Although such multivariate methods have been considered a while ago, they became really popular when a now classic study showed that patterns of brain activation could be systematically associated with perceiving pictures of different object categories.

These analyses demonstrated that there is meaningful information in the local variation in activation across voxels. In fact, the information gain when considering activation patterns is so great that reliable information can even be obtained from single trials or single brain scan images.

This greater ‘functional resolution’ power offered by multivariate analyses of neuroimaging data is very well-suited for discriminating relatively similar cognitive tasks and even makes it possible to distinguish between different mental state contents.

(Photo : Google pics)

In a study reported recently in Current Biology, used such a multivariate analysis method to ‘read out’ the location of a participant in a virtual reality environment from patterns of activation in the hippocampus, an elongated brain structure in the medial temporal lobe (shown in light grey in Figure 1). This result fits well with the existence of so-called ‘place cells’ in the hippocampus: these neurons represent particular locations in space. The hippocampus is also involved in storing spatial locations in memory: it has been shown that patterns of neural activity observed when an animal is located in a particular position in space get reactivated during sleep.

(Photo : Google pics)

Figure 1. Activation patterns in the hippocampus code for different episodic memories. The hippocampus is shown in lighter grey colour on the reconstructed medial surface of the right hemisphere of a template brain. Dashed circles represent spherical search volumes of spatially contiguous voxels which show different activation patterns depending on memory content (three short movies controlled for duration, complexity and content were used in the study). Significant but worse decoding was found in neighbouring entorhinal cortex and parahippocampal gyrus.

Decoding freely recalled memories had until now only been demonstrated for the discrimination between faces, objects and locations using activation patterns from the whole brain. In that study, the most informative parts of the activation pattern were located not in memory-associated areas but in stimulus-specific visual areas — for example, the lateral fusiform gyrus coded for the presence of faces. As reported in this issue of Current Biology, Chadwick et al. have now demonstrated successful discrimination of very similar episodic memories solely on the basis of hippocampus activation patterns (Figure 1). This suggests a functional differentiation in the representation of episodic memories within the hippocampus, similar to what is known about spatial locations, with different cell populations coding for different memory contents.